Skip to main content

Full text of "ERIC ED498458: The Distribution of Teaching and Learning Resources in California's Middle and High Schools. Issues & Answers. REL 2007-No. 023"

See other formats


ISSUES&ANSWERS 



REL 2007-No. 023 




The distribution 
of teaching 
and learning 
resources in 
California's 
middle and 
high schools 





Institute of 



NATIONAL CENTER for 
EDUCATION EVALUATION 
AND REGIONAL ASSISTANCE 

Education Sciences 



U.S. Department of Education 





ISSUES^ANSWERS 



REL 2007-No. 023 




WEST 

Regional Educational Laboratory 
AtWestEd 



The distribution of teaching and 
learning resources in California's 
middle and high schools 



September 2007 



Prepared by 

Miguel Sodas, Ph.D. 
American Institutes for Research 

Jay Chambers, Ph.D. 
American Institutes for Research 

Phil Esra 

American Institutes for Research 

Larisa Shambaugh, M.P.P. 
American Institutes for Research 



ies 



NATIONAL CENTER for 
EDUCATION EVALUATION 
AND REGIONAL ASSISTANCE 



Institute of Education Sciences 
U.S. Department of Education 




Issues & Answers is an ongoing series of reports from short-term Fast Response Projects conducted by the regional educa- 
tional laboratories on current education issues of importance at local, state, and regional levels. Fast Response Project topics 
change to reflect new issues, as identified through lab outreach and requests for assistance from policymakers and educa- 
tors at state and local levels and from communities, businesses, parents, families, and youth. All Issues & Answers reports 
meet Institute of Education Sciences standards for scientifically valid research. 

September 2007 

This report was prepared for the Institute of Education Sciences (lES) under Contract ED-06-CO-0014 by Regional Edu- 
cational Laboratory West administered by WestEd. The content of the publication does not necessarily reflect the views or 
policies of lES or the U.S. Department of Education nor does mention of trade names, commercial products, or organiza- 
tions imply endorsement by the U.S. Government. 

This report is in the public domain. While permission to reprint this publication is not necessary, it should be cited as: 

Socias, M., Chambers, J., Esra, R, & Shambaugh, L. (2007). The distribution of teaching and learning resources in California’s 
middle and high schools (Issues & Answers Report, REE 2007-No. 023). Washington, DC: U.S. Department of Education, 
Institute of Education Sciences, National Center for Education Evaluation and Regional Assistance, Regional Educational 
Laboratory West. Retrieved from http://ies.ed.gov/ncee/edlabs 



This report is available on the regional educational laboratory web site at http://ies.ed.gov/ncee/edlabs. 



Summary 

The distribution of teaching and 
learning resources in California's 
middle and high schools 



Access to important educational re- 
sources in California's middle and high 
schools is not equal among schools 
that serve different student popula- 
tions. Overall, the most disadvantaged 
populations of middle and high school 
students are likely to have the least 
access to the resources necessary for 
learning. 

Students in schools with the highest concen- 
trations of low-income students or English 
language learners are more likely to have a less 
experienced teacher or a teacher not autho- 
rized to teach that subject, and they are less 
likely to be enrolled in courses required for 
admission to the University of California (UC) 
or California State University (CSU) systems. 

Despite the general pattern of unequal dis- 
tribution of certain teaching resources, only 
limited differences in distribution are associ- 
ated with student race/ethnicity. For example, 
in foreign language and social science the 
schools with the highest concentrations of Af- 
rican American students did not experience as 
drastic a difference in out-of-field instruction 
as they have in the past. In contrast to prior 
research that indicated much more unequal 
patterns for schools with high concentrations 
of minorities, this suggests that some progress 
may have been made on one front. In addition. 



aside from the expected trend of smaller 
classes in smaller rural schools, class sizes do 
not seem to vary greatly by type of student 
population. 

The persistent disparities in the distribution 
of teaching and learning resources should not 
be ignored. Better answers are needed for why 
these patterns persist despite the time, effort, 
and resources devoted to the issue. The small 
differences identified in this report may sug- 
gest areas for further research. For example, 
to what extent can district policies and teacher 
union contracts explain the relationship be- 
tween teacher inexperience and schools with 
high concentrations of low- income students? 
Why is more out-of-field teaching found in 
schools with higher concentrations of low- 
income students? Why is much more out-of- 
field teaching found in middle schools than in 
high schools? Are fewer students enrolling in 
UC- and CSU-eligible courses in schools with 
higher concentrations of low-income students 
because the courses are not offered, because 
students are not encouraged to enroll, or 
because students do not feel prepared to take 
these courses? 

The answers to such questions will come not 
from the datasets analyzed here but from 
conversations with school, district, county, 
and state practitioners who know the realities 



iv 



SUMMARY 



of the education system. Through discussions 
with superintendents and other instructional 
leaders in the state, the underlying reasons 
behind these unequal distributions might be 
uncovered, providing local and state policy- 
makers with the information they need to 



begin to make changes. Teaching and learn- 
ing resources can, ultimately, be more equally 
distributed — the next challenge is to figure 
out how. 

September 2007 



TABLE OF CONTENTS 



Summary iii 

Main findings 1 

Teacher quality 1 

Class size 2 

What students are studying 2 

What the findings mean for future research 3 

Examining how public school systems allocate and use resources 3 

The distribution of teacher quality 5 

Out-of-field teaching 8 
Teacher education 14 
Teacher experience 18 

Average class size 21 

What students are studying 22 

University of California- and California State University- eligible classes 23 
Course-taking patterns 27 

Finding out why disparities persist 27 

Notes 28 

Appendix. Methodology 29 

References 34 

Boxes 

1 Methodology 4 

2 Key findings for out-of-field teaching 8 

3 Key findings on teacher education 14 

4 Key findings on teacher experience 19 

5 Key findings on average class size 21 

6 Key finding on eligible classes 23 

7 Key findings on course-taking 27 

Figures 

1 Percentage of students taught by out-of-field teachers, middle school, by subject and poverty quartile, 2005/06 8 

2 Percentage of students taught by out-of-field teachers, high school, by subject and poverty quartile, 2005/06 9 

3 Percentage of students taught by out-of-field teachers in University of California- and California State 

University-eligible classes, high school, by subject and poverty quartile, 2005/06 9 

4 Percentage of students taught by out-of-field teachers, middle school, by subject and English language learner 

quartile, 2005/06 10 

5 Percentage of students taught by out-of-field teachers, high school, by subject and English language learner 

quartile, 2005/06 10 

6 Percentage of students taught by out-of-field teachers. University of California- and California State 

University-eligible classes, high school, by subject and English language learner quartile, 2005/06 11 

7 Percentage of students taught by out-of-field teachers, middle school, by subject and Hispanic student quartile, 

2005/06 12 



vi 



8 Percentage of students taught by out-of-field teachers, high school, by subject and Hispanic student quartile, 

2005/06 13 

9 Percentage of students taught by out-of-field teachers, high school, by subject and population density, 2005/06 13 

10 Percentage of students taught by out-of-field teachers, middle school, by subject and school type, 2005/06 14 

11 Percentage of students taught by teacher with master’s or doctorate, middle school, by subject and poverty 

quartile, 2005/06 15 

12 Percentage of students taught by teacher with master’s or doctorate, high school, by subject and poverty 

quartile, 2005/06 15 

13 Percentage of students taught by teacher with master’s or doctorate, middle school, by subject and English 

language learner quartile, 2005/06 15 

14 Percentage of students taught by teacher with master’s or doctorate, high school, by subject and English 

language learner quartile, 2005/06 16 

15 Percentage of students taught by teacher with master’s or doctorate, middle school, by subject and Hispanic 

student quartile, 2005/06 16 

16 Percentage of students taught by teacher with master’s or doctorate, high school, by subject and Hispanic 

student quartile, 2005/06 17 

17 Percentage of students taught by teacher with master’s or doctorate, high school, by subject and African 

American student quartile, 2005/06 17 

18 Percentage of students taught by teacher with master’s or doctorate, middle school, by population density, 

2005/06 18 

19 Percentage of students taught by teacher with master’s or doctorate, high school, by population density, 

2005/06 18 

20 Percentage of students taught by teacher with master’s or doctorate, middle school, by school type, 2005/06 18 

21 Years of teaching experience, middle school, by subject and poverty quartile, 2005/06 19 

22 Years of teaching experience, high school, by subject and poverty quartile, 2005/06 20 

23 Years of teaching experience, high school, by subject and English language learner quartile, 2005/06 20 

24 Years of teaching experience, middle school, by subject and population density, 2005/06 21 

25 Years of teaching experience, middle school, by subject and school type, 2005/06 21 

26 Average class size, high school, by population density, 2005/06 22 

27 Percentage of students in University of California- and California State University- eligible classes, high school, 

by poverty quartile, 2005/06 24 

28 Percentage of students in University of California- and California State University- eligible classes, high school, 

by English language learner quartile, 2005/06 24 

29 Percentage of students in University of California- and California State University-eligible classes, high 

school, by Hispanic student quartile, 2005/06 25 

30 Percentage of students in University of California- and California State University-eligible classes, high 

school, by African American student quartile, 2005/06 26 

31 Percentage of students in University of California- and California State University- eligible classes, high school, 

by population density, 2005/06 26 

32 Percentage of students in University of California- and California State University-eligible classes, high 

school, by school type, 2005/06 26 

Tables 

1 Average teaching and learning resources in California middle and high schools, 2005/06 5 

2 Percentage of out-of-field teaching by poverty quartile and percentage of English language learners for high 

school English, 2005/06 11 

3 Admissions requirements for the University of California and California State University systems 23 

A1 Sizes of school quartiles by school characteristic 33 



MAIN FINDINGS 



1 



Access to 
important 
educational 
resources in 
California's middle 
and high schools is 
not equal among 
schools that serve 
different student 
populations. 
Overall, the most 
disadvantaged 
populations of 
middle and high 
school students 
are likely to have 
the least access 
to the resources 
necessary for 
learning. 



MAIN FINDINGS 

California policymakers are focusing on how re- 
sources are allocated and used in the public school 
system. Information on the current distribution 



of resources among schools throughout the state 
would aid their review and decisionmaking. This 
report provides a detailed and up-to-date snapshot 
of the distribution of some of the most critical re- 
sources in California’s middle and high schools — 
classroom teachers, class sizes, and college-going 
courses. It examines how these resources are 
distributed across schools grouped into quartiles 
by the percentage of low-income students, English 
language learners, Hispanic students, and African 
American students; by the population density of 
the areas in which schools are located; and by the 
type of school (traditional public or charter). 



Teacher quality 

The analysis of teacher quality focused on out- 
of-field teaching, teacher education, and teacher 
experience. Out-of-field teaching was based on 
the percentage of middle and high school students 
taking courses in specific subject areas (English, 
foreign language, math, science, and social sci- 
ence) and in special education with teachers not 
authorized to teach that subject.' Teacher educa- 
tion examined the percentage of students taught 
in these subjects by teachers with master’s degrees 
or higher. Teacher experience was measured as the 
average number of years of experience by subject. 

Out-of-field teaching. Students in schools with the 
highest concentrations of low-income students and 
those in schools with the highest concentrations of 
English language learners had a greater probabil- 
ity of experiencing out-of-field teaching, especially 
in English, math, science, and special education 
classes. Eor example, in middle school English 
classes 40.5 percent of students in first-quartile 
schools (schools with the lowest concentrations of 
low-income students) were taught by out-of-field 
teachers, compared with 48.7 percent of students 
in the fourth quartile (schools with the highest 
concentrations of low-income students).^ 

Science courses had the highest percentage of stu- 
dents receiving instruction from out-of-field teach- 
ers across all schools. English courses averaged 
a higher percentage of students with out-of-field 





2 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



teachers than did math courses. High schools, 
almost across the board, showed lower levels of 
out-of-field teaching than did middle schools, but 
the patterns with respect to poverty were similar 
to those for middle schools. High school courses 
that satisfied the University of California (UC) or 
California State University (CSU) admission re- 
quirements had lower rates of out-of-field teaching 
than all courses combined. 



across schools were analyzed. Unlike the other 
results, where clear patterns emerged with respect 
to the poverty and English language learner status 
of the students, the class-size findings were not 
striking. Few differences in average class size were 
found among the various subjects across the dif- 
ferent types of middle and high schools, except for 
rural schools. On average, classes were smaller in 
rural schools than in urban and suburban schools. 



Schools with the highest 
percentages of low- 
income students had 
the lowest percentages 
of students taught 
by teachers with 
advanced degrees 



Teacher education. Low-income 
students had less access to 
teachers with advanced degrees. 
(Advanced degrees are intended 
to measure teacher subject mat- 
ter knowledge.) Schools with the 
highest percentages of low-income 
students had the lowest percent- 
ages of students taught by teachers 
with advanced degrees. This finding held true in 
all subjects in both middle and high schools. Stu- 
dents in rural communities were also less likely to 
be taught by teachers with advanced degrees than 
were students in suburban and urban schools. 



Teacher experience. Low-income students had less 
access to experienced teachers. For example, in 
middle school English classes teachers in first- 
quartile schools (with the lowest concentrations of 
low-income students) averaged 12.4 years of expe- 
rience, while teachers in fourth- quartile schools 
(with the highest concentrations of low-income 
students) averaged 10.4 years of experience. 

Teacher experience was also higher in schools with 
lower percentages of English language learners, 
Hispanic students, and African American students 
and in schools in areas with lower population den- 
sities. Teachers in charter schools were likely to be 
much less experienced than their counterparts in 
traditional public schools. 



Class size 

Because smaller classes may afford students a 
better opportunity to learn and receive more at- 
tention from teachers, the differences in class size 



What students are studying 

Another critical component of educational success 
is access to classes that prepare students for col- 
lege. One way to assess whether different catego- 
ries of students have equal access to educational 
resources is to examine the distribution of rigor- 
ous classes and student course-taking. Another is 
to look at the distribution of enrollment in classes 
in different subjects related to college preparation 
or attendance. 

University of California- and California State 
University-eligible classes. Moving from a focus 
on teachers to a broader institutional perspective, 
we examined the rigor of courses by consider- 
ing the percentage of students enrolled in classes 
that met UC and CSU admission requirements. 
California state postsecondary school systems 
require incoming students to have taken specific 
courses in several subjects (often referred to as the 
“a-g” requirements). The different proportions of 
students enrolled in these courses may reflect dif- 
ferent levels of preparation for entry into the state’s 
public colleges. 

High schools with higher poverty levels had lower 
percentages of students enrolled in UC- and CSU- 
eligible classes, with math classes showing the 
biggest difference. 

Course-taking patterns. Another approach to 
considering the rigor of course loads in different 
schools is to look at the number of classes each 
student takes in different subjects, regardless of 
whether the courses are UC- and CSU-approved. 
For example, a heavier course load in math at a 




EXAMINING HOW PUBLIC SCHOOL SYSTEMS ALLOCATE AND USE RESOURCES 



3 



school might indicate that students there are sys- 
tematically given the opportunity or are encour- 
aged to take math instead of other courses less 
relevant to preparing for college. 

But the findings were not conclusive. Overall, there 
were no clear differences in course loads by subject 
across quartiles for the 2005/06 school year. And 
when a subject did show a clear pattern in course- 
taking, the differences between quartiles were 
generally small. 



What the findings mean for future research 

Given these patterns of unequal distribution of 
access to teaching and learning resources, future 
research should investigate the roots of these pat- 
terns to understand how best to ameliorate them. 
Researchers should also delve more deeply into 
areas that show more equal distribution of re- 
sources, to learn from successes as well as failures. 
State policymakers need to better understand how 
to create incentives for schools and districts to en- 
sure that schools with the neediest students have 
the same access to teaching and learning resources 
as other schools. 



EXAMINING HOW PUBLIC SCHOOL SYSTEMS 
ALLOCATE AND USE RESOURCES 

California policymakers are focusing on how re- 
sources are allocated and used in the public school 
system. The accountability requirements of the No 
Child Left Behind Act of 2001 have highlighted 
California’s continued low student outcomes, 
especially among the state’s large population of 
English language learners. The federal government 
also recently requested that California revise its 
plan to comply with No Child Left Behind regula- 
tions that require all teachers of core academic 
subjects to be “highly qualified,” emphasizing how 
the state plans to achieve a more equal distribution 
of highly qualified teachers throughout the state’s 
schools.^ And by the joint request of California’s 
governor, state superintendent, and senate and 
assembly leaders, an array of 22 research studies 



(under the umbrella title Getting Down to Facts) 
was released in April 2006, to provide policymak- 
ers the information needed to decide how to fund 
education and improve student outcomes (Gover- 
nor’s Committee on Education Excellence, 2006). 

Establishing whether the state’s resources are ad- 
equate for every child’s education and determin- 
ing how to address, on a state level, the ongoing 
pressure to raise student achievement outcomes 
require a better understanding of the current 
distribution of education resources throughout 
the state. This report provides a detailed snapshot 
of some of the most critical resources: middle and 
high school classroom teachers and the courses 
they teach. It examines how these resources were 
distributed in the 2005/06 school year among 
schools classified by the characteristics of their 
students (poverty status and the percentage of 
students classified as English language learner, 
Hispanic, and African American), type of school 
(public or charter), and degree of urbanization of 
the community served (measured by the popula- 
tion density of the surrounding area). (Box 1 and 
the appendix provide details on the methodology 
of the study.) The report examines how teacher 
quality (educational preparation, years of experi- 
ence, and out-of-field teaching) and class size 
were distributed across different types of schools 
and across subject areas 
within schools. It also 
presents data on the 
percentage of students 
taking courses required 
for admission to the UC 
and CSU systems (the 
“a-g” requirements). 

The report focuses on middle and high schools 
because of the current push to understand how to 
reform secondary schools, especially given the high 
dropout and low graduation rates {Education Week, 
June 2006) and the low academic achievement of 
high school students (Hall & Kennedy, 2006). The 
recent introduction of the California High School 
Exit Exam as a graduation requirement also sharp- 
ens the focus on improving secondary education. 



The report focuses on 
middle and high schools 
because of the current 
push to understand 
how to reform 
secondary schools 




4 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



BOX 1 

Methodology 

This report analyzes the distribution 
of the following education resources 
in California: 

• Out-of-field teaching. 

• Teacher education. 

• Teacher experience. 

• Class size. 

• Courses satisfying the University 
of California and California State 
University requirements. 

Each resource was analyzed in its 
relationship to the following school 
characteristics: 

• The percentage of students at the 
school who are eligible for free or 
reduced-price lunch. 

• The percentage of students at the 
school who are English language 
learners. 

• The percentage of students at the 
school who are Hispanic. 

• The percentage of students at 
the school who are African 
American. 

• The population density of the 
school’s surrounding area 
(urban, suburban, or rural). 

• Whether the school is a charter 
or a traditional public school. 

The percentages of Hispanic and Afri- 
can American students were selected 
as variables because these are the two 
most common minorities in Cali- 
fornia schools and because they are 
often most at risk within the educa- 
tion system. 



All data on middle and high schools 
were drawn from the California 
Basic Educational Data System for 
the 2005/06 school year. For each of 
the first four school characteristics 
listed above, schools were divided 
into quartiles. The first quartile was 
the group of schools with the lowest 
concentrations of students with the 
characteristic (that is, students in 
poverty or who are English language 
learners, Hispanic, or African Ameri- 
can), and the fourth quartile was the 
group of schools with the highest 
percentages of these students. The 
sizes of these quartiles are shown in 
table A1 in the appendix. In addition, 
this report presents average percent- 
ages for each quartile, but the story is 
more complex. There is quite a bit of 
variation in out-of-field teaching even 
within poverty quartiles. However, 
this analysis focuses primarily on the 
averages across quartiles. 

Next the study analyzed how the 
availability of each teaching and 
learning resource varied by percent- 
age of low-income, English language 
learner, Hispanic, and African 
American students. Students are the 
unit of analysis for this study, so all 
results are shown from the student 
perspective— that is, the percentage 
of students taught by out-of-field 
teachers, the percentage of students 
taught by teachers with advanced 
degrees, and so on. 

The graphs in this report show the 
average percentages of students 
taught by out-of-field teachers and 
teachers with advanced degrees, the 



average number of years of teaching 
experience, the average class size, and 
the average percentage of students 
enrolled in UC- and CSU-eligible 
courses for schools in each quartile. 
This report includes only the graphs 
showing the clearest patterns. (Many 
more graphs were generated for this 
study; the full results are in a sepa- 
rate technical appendix, available by 
request from the authors.) 

No tests of statistical significance were 
conducted for the differences between 
quartiles. Since the study includes 
the whole population of middle and 
high schools in California, any differ- 
ence in the distribution of education 
resources across school quartiles 
is, by definition, not an artifact of a 
particular sample. In other words, 
there is no need to test for statistical 
significance for average differences 
between quartiles, because no statisti- 
cal inference is needed. However, 
a benchmark was created defining 
small, medium-size, and large dif- 
ferences in results among quartiles, 
based on the results of the analyses. 
Any difference smaller than 0.20 stan- 
dard deviations was defined as small, 
any difference between 0.20 and 0.40 
standard deviations as medium-size, 
and any difference greater than 0.40 
standard deviations as large. 

Additional analyses were conducted 
for the two previous school years 
(2003/04 and 2004/05), but changes 
over time appeared small for many of 
the teaching and learning resources, 
so these analyses were not included 
in this report. 



THE DISTRIBUTION OF TEACHER QUALITY 



5 



TABLE 1 

Average teaching and learning resources in California middle and high schools, 2005/06 



1 Characteristic 


Middle schools 


High schools I 


Average share of students taught by out-of-field teachers (percent) 


40.4 


19.5 


Average share of teachers with advanced degrees (percent) 


32.8 


36.0 


Average teaching experience (years) 


11.8 


11.2 


Average class size (number of students) 


26.1 


25.5 


Average share of students enrolled in UC- and CSU-eligible courses (percent) 


10.1 


40.9 



Source: Authors' analysis based on data for 2005/06 from the California Basic Education Data System and other databases described in the appendix. 



With the increasing importance of a college degree 
to success in the labor market, it is necessary to 
understand the education settings that provide 
students with an adequate chance to enter college. 
For all these reasons, state policymakers need a 
more refined understanding of the distribution of 
teaching and learning resources among middle and 
high school students. 

This report does not recommend a way to equalize 
the distribution of resources, but it does highlight 
areas that others might address in bringing about 
change in California — the characteristics and 
qualifications of classroom teachers and the access 
students have to certain course offerings in middle 
and high schools. 

Table 1 lists the teaching and learning resources 
studied in this report, showing the averages for 
middle and high schools in California for the 
2005/06 school year. 

Many patterns emerge from this snapshot of 
resources, but the focus here is on the distribu- 
tion of these resources to different populations 
and schools in California. Separate analyses were 
conducted for core academic subjects — English, 
foreign language, math, science, and social 
science — as well as for special education. Descrip- 
tive statistical analyses were conducted of teaching 
personnel, class sizes, and UC- and CSU-eligible 
course-taking patterns, broken down by several 
indicators of educational need. The indicators 
include the school’s poverty level; its concentration 
of English language learners, African American 
students, and Elispanic students; its type (public 



or charter school); and the population density of 
its surrounding area. Student poverty, English 
language learner status, and minority status 
were selected as variables because they are often 
intertwined with educational success. The type of 
school was deemed important because California 
has the largest number of charter schools in the 
country (Center for Education Reform, 2006) and 
because there is interest in how charter schools 
differ from more traditional public schools. 



This report provides 
descriptive information 
about the distribution 
of teaching and learn- 
ing resources among 
California schools in 
the 2005/06 school year 
and the relationships of 
those resources to specific 
school characteristics, 
but it makes no attempt 
to imply causation. The 
intent is not to find causal 
relationships; it is to arm the state’s policymakers 
and researchers with a better understanding of the 
distribution of California’s teaching and learning 
resources— to better equip them to change the 
public education system. 



This report provides 
descriptive information 
about the distribution 
of teaching and learning 
resources among 
California schools and 
the relationships of those 
resources to specific 
school characteristics, 
but makes no attempt 
to imply causation 



THE DISTRIBUTION OF TEACHER QUALITY 

Teacher quality is among the most critical re- 
sources for a student’s education — this is univer- 
sally acknowledged. The federal government has 






6 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



recognized the importance of qualified teachers 
by requiring, through the No Child Left Behind 
Act, that all teachers of core academic subjects be 
“highly qualified.” While states have some leeway 
in defining “highly qualified,” the federal govern- 
ment did set three parameters. A highly qualified 
teacher must hold a bachelor’s degree or higher in 
the subject taught, obtain full teacher certifica- 
tion, and demonstrate knowledge in the subject 
taught. 

But measuring a teacher’s quality 
is difficult. One means of doing 
so is to determine whether the 
teacher is certified to teach the 
subject. This is sometimes a crude 
measure of quality, since uncerti- 
fied teachers may still be qualified 
and capable. For example, Kane, 
Rockoff, and Staiger (2007)— ex- 
amining high- and low-perform- 
ing teachers as defined by their students’ outcomes 
in grades 3-8 in New York City — recently found 
no difference in teaching quality between tradi- 
tionally certified teachers and alternatively certi- 
fied teachers, such as New York City’s Teaching 
Fellows and Teach for America Corps members. 
However, other research has asserted that teacher 
credentials do indeed measure teacher quality 
(Goldhaber & Brewer, 1996). 

Having a state “authorize” a teacher to teach a 
certain subject may not be a universally accept- 
able measure of teacher quality, but researchers 
do agree on the key relationship between effective 
teachers and student achievement (see, for ex- 
ample, Sanders & Rivers, 1996). So it is important 
to find proxies for teacher quality and to examine 
how these proxies are distributed among differ- 
ent schools. In the current No Child Left Behind 
policy environment the distribution of teacher 
quality — as measured by credentials, education, 
and experience — continues to be an important 
topic to study. Moreover, one could infer that 
districts value these attributes, since salary scales 
still often tie teacher compensation to certification, 
educational preparation, and experience. 



This report builds on 
previous research 
studies to present up- 
to-date information 
on the distribution of 
teaching resources 
across middle and high 
schools in California 



Whatever the measure of teacher quality, quali- 
fied teachers historically have not been equally 
distributed among schools in the United States. 
Previous studies have found that students most 
in need — those from the most disadvantaged 
communities — are often taught by the least quali- 
fied teachers (Darling-Hammond, 1987; Oakes, 
Ormseth, Bell, & Camp, 1990). The National 
Center for Education Statistics (fngersoll, 1996) 
examined the proportion and distribution of 
out-of-field teaching in the United States, based on 
the Schools and Staffing Survey from the 1990/91 
school year. Ingersoll found that schools with high 
percentages of low-income students had higher 
proportions of out-of-field teaching (defined in 
that study as teaching by someone who does not 
have at least a minor in the field taught) than did 
the schools with lower percentages of low-income 
students, but there was not as clear a pattern for 
schools with predominantly minority populations. 
In addition, the report found higher proportions 
of out-of-field teaching in middle schools than in 
high schools. 

More recently, the Public Policy Institute of 
California looked at variations in school resources 
among California schools (Betts, Rueben, & 
Danenberg, 2000). The report, using data from 
the California Basic Educational Data System for 
the 1997/98 school year, found large variations 
in teacher education and experience and in the 
percentages of teachers with full credentials. More 
experienced teachers were also typically more 
educated, so that schools with the highest concen- 
trations of low-income students had the least ex- 
perienced and least educated teachers. In addition, 
urban schools tended to have teachers with much 
less experience: teachers with master’s degrees or 
higher were most common in suburban schools, 
followed by urban schools. 

This study examines several of the same re- 
sources as that earlier report (Betts, Rueben, 8c 
Danenberg, 2000), but it uses data for the 2005/06 
school year. In addition to looking at some of the 
same resources and using an up-to-date version 
of the same database, this study shares many 




THE DISTRIBUTION OF TEACHER QUALITY 



7 



methodological characteristics with the earlier 
study. Both are focused at the school level, rather 
than at the student level, even though all results in 
this study are weighted by the number of stu- 
dents at each school. Analyses of student, class, 
and teacher characteristics in both reports are 
weighted predominantly by enrollment. However, 
the earlier study considered all K-12 students in 
California, whereas this report considers only 
middle and high school students. In addition, the 
earlier study also included multivariate regression 
results on student outcomes, whereas this report 
simply presents bivariate statistics and does not at- 
tempt to draw conclusions about the impact of the 
observed resource differences on students. 

In another recent report, the Education Trust 
(Peske & Haycock, 2006) described large gaps in 
the percentages of “highly qualified” teachers (as 
defined by No Child Left Behind) in middle and 
high schools in Ohio between schools with the 
highest and lowest percentages of low-income stu- 
dents, as well as large differences in the percent- 
ages of teachers who failed the teacher licensure 
exams in Chicago’s schools with the highest and 
lowest percentages of low-income students. In ad- 
dition, the authors reported a large gap in Illinois 
between schools with the highest and lowest mi- 
nority enrollments on the teacher quality index — a 
measure developed by the state to assess various 
attributes tied to teacher quality. A separate Edu- 
cation Trust report (Education Trust-West, 2005) 
measured the unequal distribution of teachers in 
California by examining salary differences within 
a district. While this report does not examine the 
variability of teacher salaries. Education Trust’s 
finding that high-poverty and high-minority 
schools have remarkably lower teacher salaries is 
yet another indication of an unequal distribution 
of teacher quality in California. 

More recently, a report by the Center for the Fu- 
ture of Teaching and Learning (Cuba et al, 2006) 
addressed the distribution of qualified teachers 
throughout the state. The authors found that the 
total number of “underprepared” teachers (teach- 
ers who have not completed a teacher preparation 



program and attained a preliminary or profes- 
sional credential) has declined in California over 
the past five years but that low-income and minor- 
ity students were still the most likely to have an 
underprepared teacher in the 2005/06 school year. 
The findings also indicated that the schools with 
the lowest achievement outcomes were more likely 
to have underprepared teachers. 

This report builds on these research studies and 
presents detailed information on the distribu- 
tion of teaching resources across middle and high 
schools in California, to determine whether dif- 
ferent measures of teacher qualifications show the 
same unequal distributions across schools serving 
students with varying educational needs. To gauge 
the distribution of qualified teachers throughout 
the state, it examines the distribution of out-of- 
field teaching, defined by state regulations as 
instruction by a teacher who is not authorized to 
teach the subject being taught. Previous reports 
have used other definitions of teacher quality, such 
as whether a teacher is fully credentialed, so this 
measure of out-of-field teaching may provide a dif- 
ferent assessment of the distribution of qualified 
teachers. 



Because out-of-field teaching is only one part of 
the measure of teacher quality, this report also 
examines how more experienced teachers and 
teachers with advanced degrees are distributed 
among schools. These factors are not examined 
in the recent studies of resources in California 
middle and high schools. The report explores 
variations in these three measures of teacher 
quality (out-of-field teaching, teacher education, 
and teacher experience) across schools by poverty 
quartile; percentages of 
English language learn- 
ers, African American 
students, and Hispanic 
students; type of school 
(public or charter); and 
population density of the 
surrounding area. And it 
examines how these mea- 
sures differ across subject 



Students in middle 
schools with the highest 
concentrations of low- 
income students were 
more likely to be taught 
by out-of-field teachers 
in English, math, science, 
and special education 




8 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



areas (English, foreign language, math, science, 
social science) and special education. 



Out-of-field teaching 



In both middle and 
high schools students 
in schools with the 
highest concentration 
of low-income students 
were more likely to be 
taught by out-of-field 
teachers than were 
students attending 
the schools with the 
lowest concentration 



The measure of out-of-field teaching, defined here 
as instruction by a teacher who is not authorized 
by the state to teach the subject being taught, is 
intended to capture the degree to which students 
are receiving instruction from teachers who may 
not have strong backgrounds in 
the subject and therefore may be 
less prepared to impart knowledge 
(see box 2 for key findings). 



Poverty status. In the 2005/06 
school year schools with the low- 
est percentages of low-income 
students had an average of about 
8 percent fewer students receiving 
out-of-field instruction in English, 
math, and science (a medium- 
size gap) than did schools with 



the highest percentages of low-income students. 
While foreign language, social science, and special 
education all had small gaps, the likelihood of 
out-of-field teaching increased as the concentra- 
tion of low-income students increased (figure 1). 
The percentage of students receiving out-of-field 
instruction was especially high in science classes 
in schools with the highest concentrations of low- 
income students — 62.1 percent. 

Although the percentage of students taught by 
out-of-field special education teachers in middle 
schools was lower than for the core subjects, 
schools with the highest concentrations of low- 
income students still had a higher percentage of 
students taught by out-of-field special education 
teachers than did schools with the lowest concen- 
trations of low-income students, a small gap of 3.2 
percentage points (see figure 1). But these numbers 
reflect only out-of-field teaching. When looking 
at special education by underprepared teachers 
(teachers who have not yet completed a teacher 
preparation program), others recently found the 
numbers to be much higher for special education 



BOX 2 

Key findings for out-of-field teaching 

• Students in schools with the highest concentra- 
tions of low-income students or the highest 
concentrations of English language learners had a 
greater probability of receiving instruction from 
out-of-field teachers, especially in English, math, 
science, and special education. 

• Science courses had the highest percentage of 
out-of-field teaching across all poverty quartiles. 
English courses showed, on average, a higher per- 
centage of out-of-field teaching than math. 

• High schools, almost across the board, showed 
less out-of-field teaching than did middle schools, 
but the patterns for poverty were similar to those 
shown for middle schools. 



FIGURE 1 

Percentage of students taught by out-of-field 
teachers, middle school, by subject and poverty 
quartile, 2005/06 



Quartile 


English 




Foreign 

language 




Math 




Science 




Social 

science 




Special 

education 


1st 

2nd 

3rd 

4th 



■ 6.0 
■ 6.8 
■ 8.1 
M 9.2 



■ 40.5 
H43.7 
H44.0 
^^■ 48.7 



■ 33.6 
h37.7 
38.6 

^^■ 41.7 



■ 53.7 

■ 55.5 
^■59.2 
^^■ 62.1 



■ 36.5 
h39.2 



Percent 



High school courses that satisfied the UC and CSU 
admission requirements had lower rates of out-of- 
field teaching than did all courses combined. 



Note:The first quartile has the lowest percentage of low-income students, and 
the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 




THE DISTRIBUTION OF TEACHER QUALITY 



9 



than for the other subjects (Guha et al, 2006), and 
there was a much higher percentage of under- 
prepared special education teachers in schools 
with the highest percentages of minority students 
than in schools with the lowest percentages of 
minority students. 

The average percentage of students receiving out- 
of-field teaching in math classes in 2005/06 was 
38.3 percent in middle schools but only 12.1 per- 
cent in high schools, a medium-size gap. In middle 
school English classes the average was 44.4 per- 
cent, compared with only 12.6 percent in high 
school English classes. 

The average percentage of high school students 
in classes with out-of-field teachers was lower 
in schools in the lowest poverty quartile than 
in schools in the highest poverty quartile for all 
subjects (figure 2), with medium-size gaps of 8.3 
percentage points in English and 6.1 percentage 
points in special education. As in middle schools 
the percentage of out-of-field teaching in special 
education was lower overall than for other sub- 
jects, but a medium-size gap between the first and 
fourth quartiles was still evident. 

University of California- and California State 
University-eligible courses. The results for the high 
school courses that met the UC and CSU systems’ 
admissions requirements (the “a-g” requirements) 
were similar to those for all high school courses, 
with one important difference. While the pattern 
of unequal distribution by poverty quartile was 
still evident in English and science, the average 
percentage of students exposed to out-of-field 
teaching in UC- and CSU-eligible courses (as op- 
posed to all courses) was lower across all quartiles 
and all subjects (figure 3). 

English language learner status. The distribution 
of out-of-field teaching by concentration of English 
language learners showed similar results to the 
distribution by poverty level. Schools with higher 
concentrations of English language learners had 
higher levels of out-of-field teaching in English, 
math, and science in both middle and high schools. 



FIGURE 2 

Percentage of students taught by out-of-field 
teachers, high school, by subject and poverty 
quartile, 2005/06 



Quartile 




Note: The first quartile has the lowest percentage of low-income students, and 
the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



FIGURES 

Percentage of students taught by out-of-field 
teachers in University of California- and California 
state University-eligible classes, high school, by 
subject and poverty quartile, 2005/06 



Quartile 




Note:The first quartile has the lowest percentage of low-income students, and 
the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



10 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



Special education 
students in middle 
schools with the 
highest concentrations 
of English language 
learners had a higher 
percentage of out-of- 
field teaching than 
did special education 
students in middle 
schools with the lowest 



concentrations of English 
language learners 



For middle schools the results 
for English, math, and science 
classes produced a fairly consistent 
picture. For the most part, a higher 
percentage of students received 
out-of-field teaching in schools 
with higher concentrations of Eng- 
lish language learners (figure 4). 



The highest overall proportion 
of students receiving out-of-field 
instruction in middle schools was, 
again, in science classes in schools 
with the highest concentrations of 
English language learners (62.0 per- 
cent). Special education students in middle schools 
with the highest concentrations of English language 
learners had a higher percentage of out-of-field 
teaching (9.8 percent) than did special education 
students in middle schools with the lowest concen- 
trations of English language learners (5.3 percent). 



The patterns in high school were less clear. A 
gap was seen in high school special education 



classes, in which the schools with the lowest 
English language learner concentrations had 
lower percentages of students receiving out-of- 
field teaching than those with the highest English 
language learner concentrations (figure 5). But in 
other subjects, such as foreign language, the gap 
was reversed: the schools with the lowest English 
language learner concentrations had a higher per- 
centage of students receiving out-of-field instruc- 
tion than the schools with the highest English 
language learner concentrations. 

The schools with the highest English language 
learner concentrations had higher percentages of 
students receiving out-of-field instruction in both 
English (14.8 percent) and science (42.4 percent) 
than did the schools with the lowest English lan- 
guage learner concentrations. Again, the subject 
with the highest average percentage of out-of-field 
teaching across high schools was science. 

For many subjects the percentage of out-of-field 
teaching tended to increase from the second 
to the third quartile and from the third to the 



FIGURE 4 

Percentage of students taught by out-of-field 
teachers, middle school, by subject and English 
language learner quartile, 2005/06 



FIGURES 

Percentage of students taught by out-of-field 
teachers, high school, by subject and English 
language learner quartile, 2005/06 



Quartile 


English 


1st 

2nd 

3rd 

4th 


Foreign 

language 


1st 

2nd 

3rd 


4th 


Math 


1st 

2nd 


3rd 

4th 


Science 


1st 

2nd 

3rd 

4th 


Social 


1st 

2nd 


science 


3rd 

4th 


Special 

education 


1st 

2nd 

3rd 

4th 







































h5.3 


6.6 


^^ 8.8 





55.8 

55.9 
157.3 



Percent 



Quartile 




A/ofe; The first quartile has the lowest percentage of English language learner 
students, and the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



Note:The first quartile has the lowest percentage of English language learner 
students, and the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 




THE DISTRIBUTION OF TEACHER QUALITY 



11 



TABLE 2 

Percentage of out-of-field teaching by poverty quartile and percentage of English language learners for high school 
English, 2005/06 



Poverty quartile 


English language 
learner quartile 1 


English language 
learner quartile 2 


English language 
learner quartile 3 


English language 
learner quartile 4 


Overall 

average 


First 


10.6 


7.4 


11.4 


7.8 


9.4 


Second 


11.3 


7.3 


11.1 


6.7 


9.5 


Third 


22.8 


11.2 


8.9 


12.5 


11.6 


Fourth 


30.2 


10.6 


11.4 


16.8 


17.7 


Overall average 


14.3 


8.6 


10.5 


14.8 


12.6 



Note: The first quartile has the lowest percentage of low-income or English language learner students, and the fourth quartile has the highest. 
Source: Authors' analysis based on data for 2005/06 from the California Basic Education Data System and other databases described in the appendix. 



fourth quartile, but the quartile with the lowest 
concentration of English language learners often 
had a higher percentage of out-of-field teaching 
(see figures 4 and 5). Generally, schools with high 
concentrations of low-income students can have 
low proportions of English language learners, and 
schools with high proportions of English language 
learners do not necessarily have high concentra- 
tions of low-income students. Moreover, schools 
with high concentrations of low-income students 
have more out-of-field teaching regardless of varia- 
tion in percentage of English language learners. 

For high school English classes, students in schools 
with low percentages of English language learners 
and high percentages of low-income students were 
the most likely to face out-of-field instruction in 
English, even more likely than students in schools 
with high concentrations of both English language 
learners and low-income students (table 2). This 
result was seen for other subjects as well. 

University of California- and California State 
University-eligible courses. The results were differ- 
ent for courses that met the UC and CSU systems’ 
“a-g” requirements. As with poverty status, the pro- 
portion of students receiving out-of-field teaching 
was lower across the board for UC- and CSU-eligible 
courses than for all courses. But for the majority of 
courses the pattern was opposite the patterns seen 
for all courses. Schools with the lowest proportions 
of English language learners had the highest per- 
centages of students taught by out-of-field teachers. 



FIGURE 6 

Percentage of students taught by out-of-field 
teachers. University of California- and California 
State University-eligible classes, high school, by 
subject and English language learner quartile, 
2005/06 



Quartile 




Note: The first quartile has the lowest percentage of English language learner 
students, and the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



and schools with the highest proportions of English 
language learners had the lowest percentages (fig- 
ure 6). The only exception was science, which had 
the highest percentages of out-of-field teaching. Un- 
like the other UC- and CSU-eligible courses, science 
classes had the highest percentages of students ex- 
periencing out-of-field teaching in schools with the 
highest concentrations of English language learners. 





12 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



Middle schools with the 
highest concentrations 
of Hispanic students 
had higher levels of 
out-of-field teaching in 
most subjects than did 
middle schools with the 
lowest concentrations 



with 41.2 percent of students in the 
fourth quartile receiving out-of- 
field instruction, compared with 
37.9 percent of students in the first 
quartile. This surprising pattern 
may be due in part to the fact that 
schools with the fewest English lan- 
guage learners may still have high 
concentrations of poverty. 



Minority status. For schools with different con- 
centrations of Hispanic and African American stu- 
dents the patterns were not as clear as for schools 
in the poverty and English language learner 
analyses, but there was some evidence of unequal 
distribution. 



For middle schools clear patterns emerged for 
English, math, and science classes for Hispanic 
students. In most subjects middle schools serv- 
ing higher percentages of Hispanic students had 
higher percentages of out-of-field instruction than 
schools serving lower percentages of Hispanic stu- 
dents, with small gaps in every subject (figure 7). 

For example, for math in the 2005/06 school year 
there was a 4.8 percentage point gap between the 
schools with the fewest and those with the most 
Hispanic students. The average percentage of stu- 
dents receiving out-of-field instruction in science 
was, again, higher than in any other subject. 

The patterns for African American students in 
middle schools were not as clear, but patterns 
emerged in some subjects, revealing that the 
distribution of out-of-field teachers for disadvan- 
taged populations is not always either predictable 
or unequal. 

Most high school subjects did not show a clear pat- 
tern by race or ethnicity, and the gaps between the 
first and fourth quartiles were never large. Indeed, 
the distribution was actually reversed for foreign 
language and math, with small gaps showing that 
schools with higher proportions of Hispanics have 
lower rates of out-of-field teaching. For example, in 
schools with the lowest concentrations of Hispanic 



FIGURE? 

Percentage of students taught by out-of-field 
teachers, middle school, by subject and Hispanic 
student quartile, 2005/06 



Quartile 


English 




Foreign 

language 






Math 




Science 




Social 

science 




Special 

education 


1st 

2nd 


4th 



I 7.3 
H 9.7 
h9.8 



■ 42.7 

■ 42.1 
H45.3 
^h48.0 



■ 35.7 

■ 37.1 
H39.8 
^■ 40.5 



■ 36.4 

■ 37.6 
h39.4 



■ 55.5 

■ 55.7 
^ 59.8 
^ 60.0 



Percent 



Note: The first quartile has the lowest percentage of Hispanic students, and the 
fourth has the highest. 



Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



students, 12.7 percent of students received out-of- 
field teaching in math, compared with 10.8 percent 
in schools with the highest proportions of His- 
panic students, a small gap (figure 8). 

University of California- and California State 
University-eligible courses. For high school courses 
meeting the UC and CSU “a-g” requirements, 
there were lower percentages of students in high- 
minority schools taught by out-of-field teachers 
across all quartiles than for all courses (the same 
pattern as in the poverty analysis). 

Population density. The distribution of out-of-field 
teaching across rural, suburban, and urban com- 
munities (as measured by population density) was 
analyzed. In almost every subject — English, math, 
social science, and foreign language — at both the 
middle and high school levels, rural schools had 
the highest percentage of students receiving out- 
of-field teaching (figure 9). 

The relatively high proportion of out-of-field teach- 
ing in rural schools likely reflects the difficulties 




THE DISTRIBUTION OF TEACHER QUALITY 



13 



FIGURES 

Percentage of students taught by out-of-field 
teachers, high school, by subject and Hispanic 
student quartile, 2005/06 



Quartile 




Note: The first quartile has the lowest percentage of Hispanic students, and the 
fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



FIGURED 

Percentage of students taught by out-of-field 
teachers, high school, by subject and population 
density, 2005/06 



Urban 

English Suburban 


10.5 

^H9.1 


Rural 




r . Urban 

. ^ Suburban 


Ml 

1 p' 

O' 


Urban 

Math Suburban 


^H8.9 

^H9.2 


Rural 


^^^^^21.2 










Rural 

c ■ 1 Urban 

Social , , 

Suburban 

Rural 




7.0 

^■7.4 


Special Urban 

j . Suburban 

education 


H6.4 

■ 4.4 

■ 2.7 



0 20 40 60 80 100 

Percent 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



special education teachers 
(Fideler, Foster, & 
Schwartz, 2000; AEL, 
2003). In both middle and 
high schools rural schools 
had the lowest percent- 
ages of special education 
students receiving out-of- 
field teaching. 



In every subject in 
both middle and high 
schools the percentage 
of students receiving 
out-of-field instruction 
was higher in charter 
schools than in 
traditional public schools 



Also surprising is that suburban middle and high 
schools sometimes had higher percentages of 
students taught by out-of-field teachers than did 
urban schools. This difference was large in middle 
school foreign language and in high school math 
and social science. 



University of California- and California State 
University-eligible courses. For courses that satisfy 
UC and CSU requirements the averages for out- 
of-field teaching were, again, always lower than 
for all classes, but the patterns were similar in 
all subjects except science. Rural schools had the 
highest percentages of students receiving out-of- 
field teaching in UC- and CSU-eligible courses. 

But in science suburban schools had the highest 
percentages of out-of-field teaching in UC- and 
CSU-eligible classes. 

Charter schools and traditional public schools. In 
every subject in both middle and high schools the 
percentage of students receiving out-of-field instruc- 
tion was higher in charter schools than in traditional 
public schools (figure 10). The largest gaps were 
in middle school math and science. For example, 
students in science classes in charter schools experi- 
enced 23.9 percentage points more out-of-field teach- 
ing on average than did students in public schools. 
Considering that charter schools were created to 
allow more freedom at the school level and that they 
may be less focused on teacher certification, this 
finding is not necessarily surprising. 



that rural communities have in recruiting quali- 
fied personnel in almost any specialized occupa- 
tion. But there was one exception, particularly sur- 
prising given the common shortages of certified 



University of California- and California State 
University-eligible courses. For high school UC- 
and CSU-eligible courses the percentages of stu- 
dents receiving out-of-field teaching were higher in 




14 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



FIGURE 10 

Percentage of students taught by out-of-field 
teachers, middle school, by subject and school type, 
2005/06 




Percent 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 

charter schools than in traditional public schools. 
UC- and CSU-eligible courses in both charter 
schools and public schools had lower percentages 
of out-of-field teaching than did other courses. 



Teacher education 

Another gauge of teacher quality is amount of 
teacher education. A higher level of education may 
mean that a teacher has had more exposure to the 
best practices and new methods, whether through 
a longer course of study or through returning to 
school. 

Previous research is mixed on using teacher 
education as a gauge for quality. Some studies have 
found that advanced degrees are correlated with 
higher student achievement, while others indicate 
the opposite (Greenwald, Hedges, & Laine, 1996; 
Hanushek, 1986). So while measuring teacher 
quality by whether a teacher holds an advanced 
degree is not universally accepted as the most 
effective measure of teacher quality, it is one gauge 
of teacher quality. To assess the distribution of 
teachers with advanced degrees among schools 
throughout California, the report analyzed the 



BOX 3 

Key findings on teacher education 

• Schools with the highest percentages of low- 
income students had the lowest percentages of stu- 
dents taught by teachers with advanced degrees — 
for all subjects in both middle and high schools. 

• Students in rural schools were less likely to be 
taught by teachers with advanced degrees than 
students in suburban and urban schools. 

percentages of students taught by teachers with 
master’s or doctoral degrees (box 3 summarizes 
the key findings). 

Poverty status. The general patterns across poverty 
quartiles were similar for middle and high schools. 
At both levels lower percentages of students in the 
higher poverty quartiles were taught by teachers 
with advanced degrees. In middle schools the gap 
between the first and fourth poverty quartiles 
ranged from a high of 18 percentage points for 
special education to a low of 3.9 percentage points 
for foreign language (figure 11). The pattern was 
slightly different for foreign language, but for 
English, math, science, and social science, it was 
around 10 percentage points. The gap was large for 
special education, small for foreign language, and 
medium-size for all other subjects. 

In high schools the pattern was similar (figure 12). 
Special education again had the biggest gap 
between the first and fourth poverty quartiles, 
at 16.1 percentage points, and science had the 
smallest gap, at 5.7 percentage points. The gaps in 
English, foreign language, math, and social science 
were again about 10 percentage points. The gaps 
were large for foreign language, social science, and 
special education and medium-size for the other 
subjects. 

The Public Policy Institute of California’s 2000 
report, with a focus similar to this study’s, found a 
stronger link between student poverty and teach- 
ers with lower education levels (a bachelor’s degree 
at most). It also found that schools with fewer 



THE DISTRIBUTION OF TEACHER QUALITY 



15 



FIGURE 11 

Percentage of students taught by teacher with 
master's or doctorate, middle school, by subject and 
poverty quartile, 2005/06 



FIGURE 13 

Percentage of students taught by teacher with 
master's or doctorate, middle school, by subject and 
English language learner quartile, 2005/06 



Quartile 



Quartile 



English 



Foreign 

language 



Math 



Science 



Social 

science 



Special 

education 



1st 

2nd 




English 


1st 

2nd 




4th 




4th 














3rd 




F 1 -'iy n 
language 


3rd 




4th 




4th 














3rd 




Math 


3rd 




4th 






4th 


















Science 


























2nd 




Social 


2nd 




4th 




science 


4th 








Special 










education 






4th 




4th 





0 20 40 60 80 100 0 20 40 60 80 100 



Percent 

Note: The first quartile has the lowest percentage of low-income students, and 
the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



Percent 

Note:TUe first quartile has the lowest percentage of English language learner 
students, and the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



FIGURE 12 

Percentage of students taught by teacher with 
master's or doctorate, high school, by subject and 
poverty quartile, 2005/06 



Quartile 



English 



Foreign 

language 



Math 



Science 



Social 

science 



Special 

education 



1st 

2nd 

3rd 

4th 

1st 

2nd 

3rd 

4th 

1st 

2nd 

3rd 

4th 

1st 

2nd 

3rd 

4th 

1st 

2nd 

3rd 

4th 

1st 

2nd 

3rd 

4th 
















































Percent 

Note: The first quartile has the lowest percentage of low-income students, and 
the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



low-income students had more teachers with mas- 
ter’s degrees or higher. In high schools the median 



percentage of teachers with master’s degrees or 
higher in the fourth poverty quintile was 36 per- 
cent, compared with 43 percent in the first poverty 
quintile (Betts, Rueben, & Danenberg, 2000). 

English language learner status. The patterns for 
the percentage of students taught by teachers 
with advanced degrees in schools with different 
concentrations of English language learners varied 
by subject. For the most part, students in the 
quartiles with the lowest concentrations of English 
language learners were more likely to be taught by 
teachers with advanced degrees. 

In middle schools only foreign language classes had 
a higher percentage of students taught by teachers 
with advanced degrees in schools with a higher con- 
centration of English language learners (third and 
fourth quartiles combined) than in schools with 
a lower concentration (first and second quartiles 
combined; figure 13). For all other subjects there 
were higher percentages of students taught by teach- 
ers with advanced degrees in the quartiles with 
fewer English language learners. Special education 



16 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



had the biggest gap and science the smallest. Over- 
all, the differences were smaller and less consistent 
than those for the poverty quartiles; the special edu- 
cation gap was large, but the others were small. 

In high schools the patterns were similar, except 
for foreign language classes. Unlike the distribution 
in middle schools, students in high school foreign 
language classes in schools with low concentrations 
of English language learners were more likely to be 
taught by teachers with advanced degrees (fig- 
ure 14). Special education and foreign language had 
medium-size gaps; all other gaps were small. 

Minority status. The results of an analysis of 
schools by percentage of Hispanic students were 
very similar to the results of an analysis by per- 
centage of English language learners — not surpris- 
ing considering the high percentage of English 
language learners who are Hispanic. 

In middle schools foreign language and science were 
again the only subjects with a higher percentage of 
students taught by teachers with advanced degrees 



in schools with higher concentrations of Hispanic 
students than in schools with lower concentra- 
tions of Hispanic students (figure 15). All other 
subjects had higher percentages of students taught 
by teachers with advanced degrees in the quartiles 
with lower percentages of Hispanic students. Special 
education again had the biggest gap and science the 
smallest. All gaps were small or medium-size. 

In high school the gaps were somewhat narrower. 
In percentage of students taught by teachers with 
advanced degrees, special education still showed 
the largest gap between the quartile with the low- 
est percentage of Hispanic students (46.0 percent) 
and the quartile with the highest percentage of 
Hispanic students (33.9 percent) (figure 16). This 
gap was the only one classified as large. Foreign 
language had a medium-size gap between the 
quartiles with the lowest and highest percentages 
of Hispanic students. In English, science, and 
social science the gaps were small. 

When middle schools were analyzed separately by 
percentage of African American students, no clear 



FIGURE 14 

Percentage of students taught by teacher with 
master's or doctorate, high school, by subject and 
English language learner quartile, 2005/06 



Quartile 


English 


1st 

2nd 

3rd 

4th 


Foreign 

language 


1st 

2nd 

3rd 

4th 


Math 


1st 

2nd 

3rd 

4th 


Science 


1st 

2nd 

3rd 

4th 


Social 

science 


1st 

2nd 

3rd 

4th 


Special 

education 


1st 

2nd 

3rd 

4th 


















































Percent 



FIGURE 15 

Percentage of students taught by teacher with 
master's or doctorate, middle school, by subject and 
Hispanic student quartile, 2005/06 



Quartile 
























language 












rvidiii 3 |-(j 

4th 








Science 3 rd 




4th 












4th 




— .1 *t 1 








education 





Percent 



A/ofe; The first quartile has the lowest percentage of English language learner 
students, and the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



Note:The first quartile has the lowest percentage of Hispanic students, and the 
fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



THE DISTRIBUTION OF TEACHER QUALITY 



17 



FIGURE 16 

Percentage of students taught by teacher with 
master's or doctorate, high school, by subject and 
Hispanic student quartile, 2005/06 



FIGURE 17 

Percentage of students taught by teacher with 
master's or doctorate, high school, by subject and 
African American student quartile, 2005/06 



Quartile 


English 


1st 


3rd 

4th 


Foreign 

language 


1st 

2nd 

3rd 

4th 


Math 


1st 

2nd 

3rd 




4th 


Science 


1st 

2nd 

3rd 

4th 




1st 


science 


3rd 

4th 


Special 

education 


1st 

2nd 

3rd 

4th 



■34.8 
■ 36.4 



■ 36.2 

■ 35.7 



I 32.3 
■ 36.7 



■ 33.1 

■ 32.9 

■ 35.2 
^■ 38.2 
h36.7 
h36.1 

■ 35.1 

■ 34.3 
H36.3 



Percent 



Quartile 


English 




Foreign 

language 




Math 




Science 




Social 

science 




Special 

education 




















































Percent 



Note: The first quartile has the lowest percentage of Hispanic students, and the 
fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



A/ofe:The first quartile has the lowest percentage of African American students, 
and the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



patterns were found across or within subjects. The 
high school analysis did not yield striking results 
either, except in special education, which averaged 
a notably higher proportion of students learning 
from teachers with advanced degrees in the schools 
with the fewest African Americans (figure 17). 
Conversely, the average percentage of high school 
students taught by social science teachers with 
advanced degrees was lowest in the schools with 
the fewest African American students. The gaps for 
special education and social science were medium- 
size, while those for the other subjects were small. 

Population density. The differences between urban 
and suburban schools were not large for either 
middle schools or high schools (figures 18 and 19). 
For most subjects in middle schools more students 
in suburban schools than in urban schools were 
taught by teachers with advanced degrees, but this 
pattern was reversed in high schools. More notice- 
able, though, is that at both levels there were lower 
percentages of students taught by teachers with ad- 
vanced degrees in rural schools, especially in social 
science. High schools showed a large gap between 



urban and rural schools, with 21.4 percent of rural 
students taught by teachers with advanced degrees, 
compared with 38.5 percent of urban students and 

36.2 percent of suburban students. The pattern was 
less pronounced at the middle school level, with a 
medium-size gap between urban and rural schools: 

20.2 percent of rural students were receiving 
instruction from teachers with advanced degrees, 
compared with 30.1 percent of urban students and 

33.2 percent of suburban students. 



Charter schools and traditional public schools. In 
some subjects higher percentages of students in 
charter schools than in 
traditional public schools 
were taught by teachers 
with advanced degrees, 
and in others the opposite 
was true. For the most 
part the gaps were small 
or medium-size. But in 
middle schools there was 
a large gap between the 
two types of schools for 



For most subjects in 
middle and high schools 
urban schools had the 
highest percentages 
of students taught by 
teachers with advanced 
degrees, followed by 
suburban schools and 
then rural schools 




18 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



FIGURE 18 

Percentage of students taught by teacher with 
master's or doctorate, middle school, by population 
density, 2005/06 



FIGURE 19 

Percentage of students taught by teacher with 
master's or doctorate, high school, by population 
density, 2005/06 



Urban 




English Suburban 




1 - ■ Urban 




I ^ Suburban 

language 




Urban 




Math Suburban 

Rural 




Urban 










Urban 


Social 


Rural 








Sptudl Suburban 




education 


.. . pHurarinn 



0 20 40 60 80 100 



Urban 








Urban 




Suburban 




Urban 




Suburban 




Urban 




Suburban 




Urban 




Suburban 

Rural 




Urban 




Suburban 

Rural 

{ 




20 40 



Percent Percent 

Sot/rce; Authors' analysis based on data for 2005/06 from the California Basic Source: Authors' analysis based on data for 2005/06 from the California Basic 

Education Data System and other databases described in the appendix. Education Data System and other databases described in the appendix. 



FIGURE 20 

Percentage of students taught by teacher with 
master's or doctorate, middle school, by school type, 
2005/06 



Public 

Charter 

Foreign Public 










language Charter 

Public 




Math 




Public 

Science 






Charter 








science charter 






education charter 





0 20 40 60 80 100 

Percent 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



special education: 39.8 percent of traditional pub- 
lic school special education students were taught 
by teachers with advanced degrees, compared with 
only 18.3 percent of charter school special educa- 
tion students (figure 20). 



Teacher experience 

Yet another measure of teacher quality is teacher 
experience. Studies have reported that the posi- 
tive effects of teachers on their students increase 
with years of experience (Allgood & Rice, 2002; 
Hanushek, 1986). Research also has indicated that 
the distribution of experienced teachers is not 
equal. One national study found that children in 
schools with the highest poverty concentrations 
have a “novice teacher” (three years of experience 
or less) twice as often as do children in schools 
with the lowest poverty concentrations (National 
Center for Education Statistics, 2000). The Educa- 
tion Trust (Peske & Haycock, 2006) also recently 
reported large gaps in Wisconsin in the percent- 
ages of novice (three years or less) and inexperi- 
enced (five years or less) teachers between schools 
with the highest and lowest concentrations of 
minority students and also between schools with 
highest and lowest concentrations of English 
language learners. An earlier study by the Public 
Policy Institute of California found that teachers 
with two years of experience or less were concen- 
trated in schools serving high-poverty populations 
and schools in urban areas (Betts, Rueben, & 



THE DISTRIBUTION OF TEACHER QUALITY 



19 



BOX 4 

Key findings on teacher experience 



FIGURE 21 

Years of teaching experience, middle school, by 
subject and poverty quartile, 2005/06 



• Measured by the number of years of educational 
service, teacher experience was higher in schools 
with lower percentages of low-income students; 
English language learners, Hispanic students, and 
African American students; and in schools in set- 
tings with lower population densities. 

• Teachers in charter schools were, on average, 
much less experienced than teachers in traditional 
public schools. 

Danenberg, 2000). This analysis looked at the av- 
erage number of years of experience teachers had 
across quartiles in all subjects (box 4 summarizes 
the key findings). 

Poverty status. Teachers with more experience 
were more likely to be in schools with lower con- 
centrations of low-income students. For example, 
for math classes in middle schools the difference 
between schools with the highest concentrations 
of low-income students and those with the lowest 
was more than three years of experience. Students 
in the lowest poverty quartile schools had teach- 
ers with an average of 12.9 years of experience, 
while those in schools in the highest had teachers 
with an average of 9.3 years, a large difference 
(figure 21). Students in each successive poverty 
quartile, from lowest to highest, had teachers with 
fewer average years of experience, with experi- 
ence decreasing by at least one full year in each 
quartile. A similar pattern, but with only a medi- 
um-size difference, emerged for English classes. 

The gap between schools with the highest and 
lowest concentrations of low-income students was 
less pronounced for high schools than for middle 
schools. In high schools students in the schools 
in the three lowest poverty quartiles had teach- 
ers with similar levels of experience. In math, 
English, and social science, students in schools in 
the second-lowest poverty quartile had teachers 
with the most experience. Students in the high- 
est poverty schools had teachers with the least 



Quartile 


English 




Foreign 

language 




Math 




Science 




Social 

science 




Special 

education 




















































Years of teaching experience 

Note: The first quartile has the lowest percentage of low-income students, and 
the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



experience. For example, in social science students 
in the fourth poverty quartile had teachers who 
averaged 10.8 years of experience, compared with 
12.9-13.5 years for teachers in the other three 
quartiles (figure 22). 



English language learner status. Teachers were 
also generally less experienced in schools with 
higher percentages of English language learners. 

In middle schools the difference in experience by 
percentage of English language learners was most 
pronounced in math classes. Students in schools in 
the quartile with the lowest percentages of English 
language learners had teachers who averaged 12.8 
years of experience, compared with 9.6 years for 
students in the quartile with the highest percent- 
ages, a large gap of more than three years. 



In high schools the differ- 
ence in teacher experi- 
ence between schools 
with the lowest and 
highest concentrations of 
English language learners 
was less pronounced than 



Teachers with more 
experience were 
more likely to be in 
schools with lower 
concentrations of 
low-income students 




20 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



FIGURE 22 

Years of teaching experience, high school, by subject 
and poverty quartile, 2005/06 



FIGURE 23 

Years of teaching experience, high school, by subject 
and English language learner quartile, 2005/06 



Quartile 


English 


1st 

2nd 

3rd 

4th 


Foreign 

language 


1st 

2nd 

3rd 

4th 


Math 


1st 

2nd 

3rd 

4th 


Science 


1st 

2nd 

3rd 

4th 


Social 

science 


1st 

2nd 

3rd 

4th 


Special 

education 


1st 

2nd 

3rd 

4th 



Quartile 




























































































) 5 10 15 0 5 10 

Years of teaching experience Years of teaching experience 



Note: The first quartile has the lowest percentage of low-income students, and 
the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



A/ofe: The first quartile had the lowest percentage of English language learner 
students, and the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



in middle schools. For math the medium-size gap 
between the quartiles with the fewest and most 
English language learners was only 2.2 years. The 
students in schools in the second quartile had 
teachers who were slightly more experienced than 
those in the first: 12.5 years compared with 12.3 
years (figure 23). 

Minority status. Students in the middle schools 
with the largest concentrations of African Ameri- 
can and Hispanic students typically had the 
least experienced teachers. In middle schools the 
difference in teachers’ average years of experi- 
ence between the quartiles with the lowest and 
highest percentages of Hispanic students was 2.0 
years for English (a medium-size gap) and 3.4 
years for math (a large gap), with teachers in the 
first quartile again having more experience. Math 
teachers averaged 2.1 years less experience in 
schools with the highest concentrations of African 
American students than in schools with the lowest 
concentrations, a medium-size gap. One excep- 
tion was in foreign language classes. Teachers’ 
experience was similar across all quartiles, with 
12.5 years of experience in schools in the second 



quartile and 13.2 years in schools in the other 
three quartiles. 

In high schools the patterns for students in schools 
with the highest concentrations of African Ameri- 
can students and Hispanic students were not that 
different from those in middle schools. Again, the 
gap was largest in math classes, with a large gap of 
3.5 years for math teachers between schools in the 
quartiles with the lowest and highest percentages 
of Hispanic students. Other subjects followed a 
similar pattern, though with smaller differences in 
teacher experience. 

Population density. Teachers in urban schools were 
the least experienced, those in suburban schools 
were in the middle, and those in rural schools 
were most experienced. In middle schools science 
teachers had the most variation in experience by 
population density, with 10.8 years of experience 
for urban teachers, 11.8 for suburban teachers, and 
13.4 for rural teachers — a large gap of 2.6 years 
(the gaps for all other subjects were medium-size 
or small; figure 24). The patterns were the same in 
high schools, but there was generally less variation 



AVERAGE CLASS SIZE 



21 



FIGURE 24 

Years of teaching experience, middle school, by 
subject and population density, 2005/06 



English 



Foreign 

language 



Math 



Science 



Social 

science 



Special 

education 




Years of teaching experience 



Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



FIGURE 25 

Years of teaching experience, middle school, by 
subject and school type, 2005/06 







Charter 








language Charter 




Math 1 












Charter 




science Charter 




education charter 





0 5 10 15 

Years of teaching experience 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



Teachers in charter 
schools were much 
less experienced, on 
average, than those in 
traditional public schools 



public schools. In middle 
schools there was a large 
difference in teacher ex- 
perience between charter 
schools and traditional 
public schools in all 
subjects. The biggest gap 

was in social science: teachers in traditional public 
schools averaged 5.5 more years of experience 
than those in charter schools (figure 25). Special 
education teachers were the most similar across 
the two school types, with only a 2.6-year gap. In 
high schools the experience gap was even greater 
in most subjects. 



AVERAGE CLASS SIZE 

Aside from the type of teacher a student has, 
an important factor in an effective educational 
environment is the number of students learning 
together in each classroom (hox 5 summarizes the 
key findings on class size). The Tennessee Student/ 
Teacher Achievement Ratio (STAR) experiment, 
one of the first randomized control trials to 
examine class size in education, found signifi- 
cant and positive effects for smaller classrooms 
(Finn & Achilles, 1999; Nye & Hedges, 1999). But 
more recent work hy Milesi and Gamoran (2006) 
found no positive effect for reducing class size in 
kindergarten. 

Class size has been the focus of much attention 
nationwide and in California. The Morgan-Hart 
Class Size Reduction Act of 1989 encourages 



BOX 5 

Key findings on average class size 



across categories. The experience for science teach- 
ers ranged from 11.5 years for urban teachers to 
12.9 years for rural teachers — a medium-size gap. 

Charter schools and traditional public schools. 
Teachers in charter schools were much less ex- 
perienced, on average, than those in traditional 



• Few differences were found in average class size 
for the various subjects across the different types 
of schools. 

• Average class sizes were smaller in suburban 
schools than in urban schools, and smaller again 
in rural schools than in suburban schools. 




22 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



districts to reduce class size in ninth-grade Eng- 
lish and in one other ninth-grade course required 
for graduation. The Class Size Reduction Act of 
1996, focused solely on kindergarten through 
third grade, stirred up much debate in California 
about the usefulness of smaller class sizes. As a 
large proportion of the governor’s recent $2.9 bil- 
lion settlement with the teachers union is allocated 
to reducing class size again in California through 
the Quality Education Investment Act (SB 1133), a 
better understanding of the current distribution of 
class size in the state is important. 

Previous work by the Public Policy Institute of 
California (Betts, Rueben, &Danenberg, 2000) 
that examined variations in school resources 
based on the proportions of disadvantaged popula- 
tions in California schools did not find a large dif- 
ference in average class size for the 1997/98 school 
year. This report finds similar results in California 
middle and high schools for 2005/06. The only 
clear difference in average class size was found, 
not surprisingly, among schools in settings with 
different population densities — urban, suburban, 
and rural (figure 26). 



FIGURE 26 

Average class size, high school, by population 
density, 2005/06 



English 



Foreign 

language 



Math 



Science 



Social 

science 



Special 

education 



Urban 








Urban 






Rural 

Urban 

Suburban 










Urban 






Rural 

Urban 








Rural 






Urban 




Suburban 




Rural 





0 10 20 30 40 50 



Average class size 



Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



average than did suburban schools, and suburban 
schools had larger class sizes on average than did 
rural schools in all subjects except middle school 
special education and foreign language. 



Classes in urban schools 



were larger on average 
than those in suburban 
schools, and classes 
in suburban schools 
were larger than 
those in rural schools, 
but the differences 
were never large 



The average class sizes experienced by students in 
schools by poverty, English language learners, and 
minorities show small to no differences in average 
class size. The small differences are almost always 
due to slightly smaller classes in the schools with 
the lowest concentrations of poverty, English 
language learners, or minorities. For example, for 
high school English the schools with the low- 
est proportions of Hispanic students had almost 
two fewer students per class on average than did 
schools with the highest propor- 
tions of Hispanic students (24.5 
students and 26.2 students). 



One clear pattern was the differ- 
ence in average class size across 
schools in urban, suburban, and 
rural communities. But the dif- 
ferences between urban and rural 
schools were never large. Urban 
schools had larger class sizes on 



WHAT STUDENTS ARE STUDYING 

Another critical component of educational suc- 
cess in California is access to classes that prepare 
students for college. One way to assess whether 
different categories of students have equal access 
to educational resources is to examine the dis- 
tribution of rigorous classes and student course- 
taking. For example, higher enrollment in more 
challenging classes among one group of students 
may indicate that that group is better prepared for 
the future than another. 

Admissions requirements for the UC and CSU 
systems include at least 15 year-long classes in a 
variety of subjects, at least 7 of them in the last 
two years of high school. For a class to meet the 
requirements, a school must submit an annual 
request to the University of California and have 
the course certified by the university (table 3). 

The California State University system uses the 




WHAT STUDENTS ARE STUDYING 



23 



TABLE 3 

Admissions requirements for the University of 
California and California State University systems 



1 Subject 


Years required 


History/social science 


2 


English 


4 


Math 


3, with 4 recommended 


Laboratory science 


2, with 3 recommended 


Language other 
than English 


2, with 3 recommended 


Visual and performing arts 


1 


College preparatory 
electives 


1 



Source; www.universityofcalifornia.edu/admissions/undergrad_adm/ 
paths_to_adm/freshman/subject_reqs.htmi. 



University of California’s list of approved courses. 
The UC- and CSU-required coursework is designed 
to ensure that incoming students have “attained 
a body of general knowledge that will provide 
breadth and perspective to new, more advanced 
study” and have “attained essential critical think- 
ing and study skills.”'^ 

A 2000 report by the Public Policy Institute 
of California examined variations in school 
resources among California schools and found, 
for the 1997/98 school year, that the median 
percentage of classes offered that met UC and 
CSU requirements ranged from about 50 percent 
for the two highest poverty quintiles to about 
60 percent for the lowest poverty quintile (Betts, 
Rueben, 8c Danenberg, 2000). The study also 
reported that disadvantaged students received 
fewer resources in individual districts — not just 
in different districts across the state. Suburban 
schools were most likely to offer UC- and CSU- 
eligible classes, followed by urban and then rural 
schools. 

The University of California All Campus Consor- 
tium on Research for Diversity and the University 
of California at Los Angeles Institute for Democ- 
racy, Education, and Access published a report 
(Oakes et al, 2006) that examined questions 
similar to those addressed here. Using data from 



BOX 6 

Key finding on eligible classes 

• Higher-poverty schools had lower percentages of 
students enrolled in UC- and CSU-eligible classes 
in high schools, with math showing the biggest 
difference. 

the 2004/05 school year, the report found that 
schools with high proportions of African Ameri- 
can and Hispanic students had lower numbers of 
UC- and CSU-eligible courses than schools with 
majority white or Asian students, leading to lower 
completion rates of UC and CSU requirements in 
high-minority schools. 

This section builds on the work done by those 
studies (Betts, Rueben, & Danenberg, 2000; 
Oakes at al., 2006) by looking at the most cur- 
rent data available for the 2005/06 school year. It 
examines the course-taking patterns in differ- 
ent subjects and the distribution of enrollment 
in classes that meet UC and CSU requirements 
in different subjects (box 6 summarizes the 
key finding). Because very low percentages of 
students are enrolled in UC- and CSU-eligible 
courses in middle schools, this analysis consid- 
ers only high school students. Also excluded 
are foreign language classes (because almost all 
foreign language classes at the high school level 
are UC- and CSU-eligible) and special education 
classes (because no special education classes are 
UC- and CSU-eligible). 



University of California- and California 
State University-eligible classes 

Poverty status. In high schools, schools with 
higher concentrations of low-income students had 
fewer students enrolled in UC- and CSU-eligible 
classes in math, English, and social science. The 
biggest difference was in math: 44 percent of math 
students in schools with the lowest poverty con- 
centrations were in UC- and CSU-eligible classes, 
compared with only 25.5 percent in schools with 
the highest poverty concentrations— a large gap 





24 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



FIGURE 27 

Percentage of students in University of California- 
and California State University-eligible classes, high 
school, by poverty quartile, 2005/06 



FIGURE 28 

Percentage of students in University of California- 
and California State University-eligible classes, high 
school, by English language learner quartile, 2005/06 



Quartile 



Quartile 



English 



Math 



Science 



Social 

science 



1st 

2nd 

3rd 

4th 

1st 

2nd 

3rd 

4th 

1st 

2nd 

3rd 

4th 

1st 

2nd 

3rd 

4th 



0 



^■ 44.0 

37.4 



■ 73.5 

172.8 
71.3 

171.9 



20 40 60 80 100 




English 



Math 



Science 



Social 

science 




Percent 



Percent 



Note: The first quartile has the lowest percentage of low-income students, and 
the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



Note:The first quartile has the lowest percentage of English language learner 
students, and the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



(figure 27). In English the difference was only 
3 percentage points (17.4 percent compared with 
14.6 percent, a small gap). In social science the 
difference was 10 percentage points (25.5 percent 
compared with 15.1 percent, a medium-size gap). 



English language learner status. Overall, the 
patterns of enrollment in UC- and CSU-eligible 
courses for schools with different percentages of 
English language learners were somewhat similar 
to those in the poverty analysis, with one notice- 



able difference. 



In high schools, 
schools with higher 
concentrations of 
low-income students 
had fewer students 
enrolled in University of 
California- and California 
State University-eligible 
classes in math, English, 
and social science 



There were higher percentages 
of students in eligible classes 
in the second English language 
learner quartile than in the first 
quartile for math, English, sci- 
ence, and social science classes 
(figure 28). Eor example, among 
math students 41.6 percent in 
schools in the second English 
language learner quartile were 
in math classes meeting UC and 
CSU requirements compared with 
37.7 percent in the first quartile, a 



3.9 percentage point gap. This pattern is similar to 
that for out-of-field teaching. 

One possible explanation for this pattern is that, 
as previously mentioned, out-of-field teaching 
increased drastically with poverty for schools 
in the first English language learner quartile. 
Students in schools with low percentages of 
English language learners and high percentages of 
low-income students were very likely to face out- 
of-field instruction in English — even more likely 
than students in schools with high concentrations 
of English language learners and of low-income 
students. Similar results in this section also seem 
to imply that the interaction between poverty and 
English language learners can partially explain 
why the average enrollment in UC- and CSU-eli- 
gible classes is lower for the first English language 
learner quartile. 

Minority status. The patterns for minority stu- 
dents’ enrollment in UC- and CSU-eligible courses 
were not clear. To some extent, there was higher 
enrollment in such courses in schools with lower 




WHAT STUDENTS ARE STUDYING 



25 



concentrations of Hispanic students, but the gap 
between quartiles was smaller than that between 
the lowest and second-lowest English language 
learner quartiles. 

With Spanish-speaking students making up 
about 83 percent of California’s English language 
learner student population,^ it is not surprising 
that the results for Hispanic students are very 
similar to those for English language learners. But 
in math the lowest and second-lowest quartiles 
for Hispanic students follow the same pattern as 
the other quartiles (figure 29 shows this for high 
school), instead of being reversed as they are for 
English language learners. In the first quartile 
42.7 percent of students are in UC- and CSU- 
eligible classes. This proportion drops for each 
quartile as the percentage of Hispanic students 
increases, down to 26.0 percent for the fourth 
quartile. The gap between the first and fourth 
quartiles is large. 

In English the gap between the first and fourth 
quartiles of Hispanic students is slightly more 
than 2 percentage points, with fewer students in 
the fourth quartile taking UC- and CSU-eligible 
classes (14.4 percent compared with 16.7 percent, 
a small gap). At 17.1 percent the percentage in the 
second quartile of Hispanic students is slightly 
higher than the percentage in the first, but the 
overall pattern is one of fewer students taking 
classes meeting UC and CSU requirements in 
quartiles with more Hispanic students. 

Social science follows the same general pattern, 
with 23.1 percent of students taking classes that 
meet UC and CSU requirements in schools in 
the first quartile of Hispanic students, just over 
21 percent in the second and third quartiles, and 
only 15.8 percent in the fourth. The gap between 
the first and fourth quartiles of Hispanic students 
is small. 

For schools with varying concentrations of 
African American students the differences 
between quartiles were smaller than for schools 
with varying concentrations of Hispanic 



FIGURE 29 

Percentage of students in University of California- 
and California State University-eligible classes, high 
school, by Hispanic student quartile, 2005/06 

Quartile 




Percent 

Note:The first quartile has the lowest percentage of Hispanic students, and the 
fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 

students — and the patterns were less distinct. 

In math the two quartiles with lower concentra- 
tions of African American students had higher 
percentages of students enrolled in UC- and 
CSU-eligible classes than the two quartiles with 
higher concentrations of African American stu- 
dents (34.4 percent and 38.5 percent, compared 
with 34.0 percent and 31.4 percent). In science, 
however, there was a small difference between 
the percentage of students enrolled in eligible 
classes in the quartile with the lowest percent- 
age of African American students and the 
percentages in the other quartiles: 69.5 percent 
for the first quartile, compared with between 
73.6 percent and 74.2 percent for the other three 
quartiles (figure 30). 

Population density. Rural high schools gener- 
ally had the fewest students in UC- and CSU- 
eligible classes, while urban schools generally had 
the most. 

In both math and social science urban and subur- 
ban schools had similar percentages of students 



26 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



FIGURE 30 

Percentage of students in University of California- 
and California State University-eligible classes, high 
school, by African American student quartile, 2005/06 

Quartile 



English 



Math 



Science 



Social 

science 




Percent 



FIGURE 32 

Percentage of students in University of California- 
and California State University-eligible classes, high 
school, by school type, 2005/06 



English 



Math 



Science 



Social 

science 




Percent 



Note: The first quartile has the lowest percentage of African American students, 
and the fourth has the highest. 

Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 



FIGURE 31 

Percentage of students in University of California- 
and California State University-eligible classes, high 
school, by population density, 2005/06 




Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 

in classes that met UC and CSU requirements 
(figure 31). Rural schools had fewer students in 
eligible classes in both subjects. 



Source: Authors' analysis based on data for 2005/06 from the California Basic 
Education Data System and other databases described in the appendix. 

In English, suburban schools were more like rural 
schools than like urban schools, and suburban and 
rural schools had lower percentages of students 
in UC- and CSU-eligible classes than did urban 
schools. Rural schools had 14.5 percent of their 
English students in eligible classes, suburban 
schools had almost the same (14.3 percent), and 
urban schools had 18.2 percent. 

In science the percentage of students in suburban 
schools taking classes that met the requirements 
fell neatly between the percentages for rural and 
urban schools. Rural schools were at 63.3 per- 
cent, suburban schools at 71.2 percent, and urban 
schools at 76.6 percent. The gap between rural and 
urban schools was medium-size. 

Charter schools and traditional public schools. In 
high schools across all subjects there were lower 
percentages of charter school students than public 
school students in UC- and CSU-eligible classes. 
The differences ranged from a medium-size gap of 
12.7 percentage points fewer charter students in 
science classes to a small gap of 3.1 points fewer 
in English. In both types of schools science had 
the highest percentage of students in classes that 



FINDING OUT WHY DISPARITIES PERSIST 



27 



met the UC and CSU requirements: 73.7 percent 
for traditional public schools and 61.0 percent for 
charter schools (figure 32). 



Course-taking patterns 

The analysis also looked at the distribution of 
enrollment in classes in different subjects re- 
lated to college preparation or attendance. A 
heavier course load in math for the students in 
one quartile, for example, might indicate that 
students in that quartile are systematically given 
the opportunity, or encouraged, to take math 
instead of other courses less relevant to preparing 
for college (box 7 summarizes the key findings on 
course-taking). 

Across poverty quartiles in middle schools math 
showed the clearest pattern. Students in schools 
with higher concentrations of low-income students 
averaged slightly heavier math course loads than 
students in schools with lower concentrations of 
low-income students. The range was from an aver- 
age of 0.95 class periods for students in the first 
poverty quartile to an average of 1.03 class periods 
for students in the fourth. 

Conversely, the pattern for middle school science 
shows that students in schools with the lowest 
concentrations of low-income students were taking 
an average of 0.91 periods of science, compared 
with 0.80 periods for students in schools with the 
highest concentrations, with incrementally lower 
course loads in each quartile from lowest poverty 
to highest. In high schools English showed the 

BOX? 

Key findings on course-taking 

• Overall, there were no clear differences in course 
loads by subject across quartiles for the 2005/06 
school year. 

• When a subject did show a clear pattern in course- 
taking, the differences between quartiles were 
generally small. 



clearest pattern in the poverty analysis. Stu- 
dents in schools with the lowest concentrations 
of low-income students were taking an average 
of 1.03 periods of English, compared with 1.19 
periods for students in schools with the highest 
concentrations. 



FINDING OUT WHY DISPARITIES PERSIST 

Many of the findings in this report are not new. 
Many of the unequal patterns found in 2005/06 
are similar to patterns that other researchers have 
found over the past 30 years. Out-of-field teach- 
ing is still distributed unequally — occurring less 
frequently in schools with the lowest proportions 
of disadvantaged students and English language 
learners. Teachers with the most experience con- 
tinue to teach at schools with the lowest concentra- 
tions of low-income students and English language 
learners. Enrollments in college-preparation 
courses that allow students to enter California’s 
state college and university systems are lowest in 
schools where poverty is highest. While it is not 
surprising that more experienced teachers gravi- 
tate to less disadvantaged schools, it is discourag- 
ing to see how little this trend has changed. De- 
spite an ongoing, overarching focus on educational 
success for all children, little change is evident in 
the distribution of these resources. 

Within the general patterns of unequal distribu- 
tion of certain teaching resources, however, only 
limited differences associated with students’ 
race or ethnicity were observed. Eor example, in 
foreign language and social science classes the 
schools with the highest concentrations of African 
American students did not experience as drastic 
a difference in out-of-field instruction as other 
at-risk populations. Contrary to previous research 
that indicated much more unequal patterns for 
schools with high concentrations of minorities, 
this suggests that some progress may have been 
made on one front. 

The persistent disparities in the distribution 
of teaching and learning resources should not 



28 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



The persistent disparities 
in the distribution of 
teaching and learning 
resources should 
not be ignored 



be ignored. Better answers are 
needed for why these patterns per- 
sist despite the expenditure of so 
much time, effort, and resources. 
Finding answers within the small 
differences in this report may sug- 
gest areas for further research. For 
example, to what extent can district policies and 
teacher union contracts explain the relationship 
between teacher inexperience and schools with 
high concentrations of low-income students? Why 
is there more out-of-field teaching in schools with 
higher concentrations of low-income students? 
Why is there dramatically more out-of-field teach- 
ing in middle schools than in high schools? Are 
fewer students enrolling in the UC- and CSU-eligi- 
ble courses in schools with higher concentrations 
of low-income students because the courses are 
not offered, because students are not encouraged 
to enroll, or because students do not feel prepared 
to take these courses? 



The answers to such questions can be found not in 
the datasets analyzed here, but in conversations 
with school, district, county, and state practi- 
tioners who know the realities of the education 
system. Through discussions with superintendents 
and other instructional leaders in the state, the 
underlying reasons behind these unequal dis- 
tributions might be uncovered, providing local 
and state policymakers with the information 
they need to begin to make changes. Teaching 
and learning resources can, ultimately, be more 
equally distributed — the next challenge is to 
figure out how. 

An area of further research would involve apply- 
ing the standard behavioral models generally used 
by economists to explore the range of factors that 
affect the choices of school and district decision- 
makers. Next steps in this work should build on 



discussions with administrators at the school 
and district levels and then proceed to develop- 
ing models that can help us explore the various 
relationships among state funding, local policies 
about resource allocation, and other regional 
characteristics that influence the labor market in 
school districts. 

These additional pieces of research can help in 
ascertaining how much certain policies and 
behaviors might reduce the disparities that persist 
for students with the greatest needs. 



NOTES 

1. According to this definition, an out-of-field 
teacher may or may not be fully credentialed 
but does not have state authorization to teach 
the subject he or she is currently teaching. 

2. This report presents average percentages for 
each quartile, but the story is more complex. 
Out-of-field teaching varies even within pov- 
erty quartiles. However, this analysis focuses 
primarily on the averages across quartiles. 

3. The definition of a highly qualified teacher is 
determined in part by each state, but the fol- 
lowing are requirements: a bachelor’s degree 
or higher in the subject taught, full teacher 
certification, and demonstrated knowledge in 
the subject taught. 

4. From www.universityofcalifornia.edu/ 
admissions/undergrad_adm/paths_to_adm/ 
freshman/subj ect_reqs .html 

5. California Department of Education, Califor- 
nia Basic Educational Data System (CBEDS) 
and Language Census Data Eiles (R30-LC). 




APPENDIX 



29 



APPENDIX 

METHODOLOGY 

This appendix explains in further detail how 
the analyses in this report were performed and 
discusses the data sources used for the analyses. It 
also shows the number of schools in each quartile, 
population density group, and school type (see 
table A1 at the end of this appendix). 



Teaching and learning resources 

Out-of-field teaching. Teacher-level information was 
taken from the Professional Assignment Informa- 
tion Form (PAIF) data in the California Basic Edu- 
cational Data System (CBEDS) file for the 2005/06 
school year. This staff characteristics file lists 
teachers by ID variable and County-District-School 
code and is publicly available for download from 
the California Department of Education web site 
(www.cde.ca.gov/ds/ss/cb/filespaifasp). The layout 
of the database for the school year 2005/06 can be 
found at http://www.cde.ca.gov/ds/ss/cb/fspaif05. 
asp. The PAIF staffing files indicate which subjects 
each teacher is authorized to teach in California. 

For instance, the math variable in that database 
has a value of one if a teacher is authorized to teach 
that subject in the state and zero if not. 

The PAIF staffing files were merged (by record iden- 
tification and County-District-School code) with 
the PAIF course data. This course database (see the 
layout of the 2005/06 database at http://www.cde. 
ca.gOv/ds/ss/cb/fsassign05.asp) contains informa- 
tion about all the courses that have been taught in 
California schools, by grade and subject. Merging 
these two databases allows the analysis of which 
teachers taught which courses in the state. Note 
that the PAIF course database contains a variable 
identifying the code of the assignment. In order to 
understand what these codes really mean, it is nec- 
essary to merge (using the assignment code) a third 
database. This third database, the assignment code 
and name file (see the layout of this file at http:// 
www.cde.ca.gov/ds/ss/cb/fsasgncode.asp), makes 
it easy to analyze whether a teacher is teaching a 
course that he or she is authorized to teach. 



As a second step, it is necessary to aggregate all 
the teacher- and course-level information and to 
generate a school-level percentage of, for example, 
math courses taught by nonauthorized (out-of- 
field) teachers. In this school-level computation 
each course’s teacher authorization was weighted 
by its total enrollment (which comes from the 
PAIF course data file). Therefore, the school-level 
percentage of courses of subject i that are taught 
by nonauthorized teachers is equal to: 



School level 
% out of field / 
teaching. = 1 1 “ 



N 

zl Enrollment., x 



Authorization 
dummy, variable 



N 

y Enrollment.. 



X 100 



where; denotes the different instances of a specific 
assignment or subject at the school. 

Once these school-level percentages have been 
estimated, poverty, English language learner, His- 
panic, and African American quartile averages, as 
well as charter, traditional public, urban, subur- 
ban, and rural school averages, are calculated by 
taking the simple mean of all the schools that fall 
into each group. 

University of California and California State 
University course requirements. The assignment 
code and assignment name files (http://www.cde. 
ca.gov/ds/ss/cb/fsasgncode.asp) contain data that 
indicate whether each course taught in California 
schools is eligible to be designated as meeting the 
University of California (UC) and California State 
University (CSU) requirements for admission. The 
four possible values of this variable are: always 
or almost always approved as UC “a-g” course, 
cannot be approved as UC “a-g” course, sometimes 
approved as UC “a-g” course, and not a teaching 
assignment. With this information, the percentage 
of all the courses in each school that are always 
or almost always approved as UC “a-g” courses 
was computed. (Courses that were sometimes 
approved were not included, since this variable 
does not indicate whether, in the end, they were 
actually approved. Courses that were classified as 
nonteaching assignments were also not included 



30 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



in the analysis.) The school-level variable is then 
defined as 

N 

School level % S Enrollment, x UC/CSU dummy variable. 

UC/CSU courses. = — x 100 

' N 

y Enrollment.. 

where i denotes the subject and j the different 
instances of a specific assignment or subject at the 
school. The UC/CSU dummy variable takes a value 
of one if the course satisfies UC and CSU require- 
ments and zero otherwise. 

Once these school-level percentages have been 
estimated, poverty, English language learner, 
Hispanic, and African American quartile aver- 
ages, as well as charter, traditional public, urban, 
suburban, and rural school averages, are easily 
calculated by taking the simple mean of all the 
schools that fall into each group. 

Class size. The class size variable is a measure of 
the enrollment of each course. This enrollment 
figure is estimated from the course database 
(http://www.cde.ca.gov/ds/ss/cb/fsassign05. 
asp) by adding the male and female enrollments 
of the class or course. Note that this estimated 
total enrollment of each course is also used as a 
weight in the school-level percentages of out-of- 
field teaching and classes satisfying UC and CSU 
requirements. 

N 

y Enrollment.. 

School level average class size. = ■ ^ x 100 

where i denotes the subject,; the different in- 
stances of a specific assignment or subject at the 
school, and N the total number of instances of a 
specific assignment. 

Once these school-level average class sizes have 
been estimated, poverty, English language learner, 
Hispanic, and African American quartile averages, 
as well as charter, traditional public, urban, sub- 
urban, and rural school averages, are calculated by 
taking the simple mean of all the schools that fall 
into each group. 



Course-taking patterns. One of the analyses exam- 
ines the percentage of students in each school who 
enrolled in a particular subject. To generate these 
data it is necessary to use two different sources. 

The first one is the PAIE course database (www. 
cde.ca.gOv/ds/ss/cb/fsassign05. asp), which lists the 
total enrollment in each course. The second is the 
CBEDS School Information Form (for a layout of 
this data file see http://dq.cde.ca.gov/DataQuest/ 
downloads/fs/sifenr.asp?year=2005), which pro- 
vides total school enrollment data. The percentage 
of students at each school who enroll in a particu- 
lar subject is then estimated as follows: 

N 

y Enrollment.. 

School level % of students enrolled in subject. = -T, . , x 100 

' Total enrollment 

where i denotes the subject and; the different 
instances of a specific assignment or subject at 
the school. The denominator indicates the total 
student enrollment at the school as taken from the 
CBEDS database. 

Once these school-level average class sizes have 
been estimated, poverty, English language learner, 
Hispanic, and African American quartile averages, 
as well as charter, traditional public, urban, sub- 
urban, and rural school averages, are calculated by 
taking the simple mean of all the schools that fall 
into each group. 

Teacher experience. The professional experience of 
teachers is taken from the PAIE staff files (www. 
cde.ca.gOv/ds/ss/cb/fspaif05.asp). A variable in that 
file indicates the total number of years of public 
and private educational service in the current 
school district, other districts, other states, and 
other countries. 

The analysis considers teacher experience across 
different subjects, so it is necessary to use the PAIE 
staff file information on teaching experience in 
conjunction with the PAIE assignment database 
(www.cde.ca.gov/ds/ss/cb/fsassign05.asp) to match 
teachers to subjects. Again, this merging is done 
using the teacher ID variable and County-District- 
School code. 



APPENDIX 



31 



Once these two data files have been merged, the 
school-level average teacher experience is esti- 
mated as follows: 

N 

Average years of E Teacher experience,^ Enrollment,^ 
teacher experience, = ■ ^ x ^ x 100 

' y Enrollment.. 

where i denotes the subject,; denotes the different 
instances of a specific assignment or subject at the 
school, and T denotes the number of teachers at 
the school who teach a particular subject. 

Once these school-level averages have been esti- 
mated, poverty, English language learner, His- 
panic, and African American quartile averages, as 
well as charter, traditional public, urban, subur- 
ban, and rural school averages, are calculated by 
taking the simple mean of all the schools that fall 
into each group. 

Teacher education. The PAIF staff database (www. 
cde.ca.gOv/ds/ss/cb/fspaif05. asp) contains a vari- 
able that lists the educational attainment of each 
teacher in California. There are six possible values 
of this variable: 

1 = Doctorate 

2 = Master’s degree plus 30 or more semester 

hours 

3 = Master’s degree 

4 = Bachelor’s degree plus 30 or more semester 

hours 

5 = Bachelor’s degree 

6 = Less than bachelor’s degree 

Since the focus of the analysis is the percentage of 
students taught by teachers with master’s or doc- 
toral degrees, the first step is to generate a dummy 
variable that takes a value of one if the teacher’s 
education variable is 1, 2, or 3. If the teacher’s edu- 
cation variable is 4, 5, or 6, the dummy variable 
takes a value of zero. 

As with the analysis of teacher experience, to esti- 
mate figures for teacher education by subject, it is 
necessary to merge the PAIF staff data file with the 



respective assignment database (http://www.cde. 
ca.gOv/ds/ss/cb/fsassign05. asp). Knowing which 
teacher is assigned to each subject, the percentage 
of students being taught by teachers with ad- 
vanced academic degrees is computed as follows: 

% of students 
taught by ^ 

teachers with 2 Advanced degree dummy variable, x Enrollment 

advanced degrees. = ^ ^ ^ x 100 

y Enrollment. 

where i denotes the subject and; the different 
instances of a specific assignment or subject at the 
school. 

Once these percentages have been estimated, 
poverty, English language learner, Hispanic, and 
African American quartile averages, as well as char- 
ter, traditional public, urban, suburban, and rural 
school averages, are calculated by taking the simple 
mean of all the schools that fall into each group. 



School characteristics 

Percentage of students eligible for free or reduced- 
price lunch. This variable is generated using 
two different data sources. First, the number of 
students eligible for free or reduced-price lunch 
comes from the school-level data of the Free and 
Reduced Price Meals Program and the Califor- 
nia Work Opportunity file for 2005 (see layout at 
www.cde.ca.gov/ds/sh/cw/fsafdc2.asp). The data 
file can be downloaded from www.cde.ca.gov/ 
ds/sh/cw/filesafdc.asp. The denominator of this 
ratio is obtained from the CBEDS School Infor- 
mation Form (for a layout of this data file, see 
http://dq.cde.ca.gov/DataQuest/downloads/fs/ 
sifenr.asp?year=2005), which provides total school 
enrollment data. 

Percentage of Hispanic students. This percentage 
was taken directly from the school-level academic 
performance index (API) 2005 database, which 
can be downloaded from www.cde.ca.gov/ta/ac/ 
ap/apidatafiles.asp. For a description of the layout 
of the database, see www.cde.ca.gov/ta/ac/ap/ 
reclayout05b.asp. 



32 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



Percentage of English language learners. This 
percentage was taken directly from the school- 
level API 2005 database, which can be downloaded 
from www.cde.ca.gov/ta/ac/ap/apidatafiles.asp. 

For a description of the layout of the database, see 
www.cde.ca.gov/ta/ac/ap/reclayout05b.asp. 

Percentage of African American students. This 
percentage was taken directly from the school- 
level API 2005 database, which can be downloaded 
from www.cde.ca.gov/ta/ac/ap/apidatafiles.asp. 

For a description of the layout of the database, see 
www.cde.ca.gov/ta/ac/ap/reclayout05b.asp. 

Charter school. The value of this variable was de- 
termined by consulting the school-level API 2005 
database, which can be downloaded from www. 
cde.ca.gov/ta/ac/ap/apidatafiles.asp. For a descrip- 
tion of the layout of the database, see www.cde. 
ca.gOv/ta/ac/ap/reclayout05b.asp. 

Population density. The value of this variable 
was determined by consulting the CBEDS public 
schools database (see layout at www.cde.ca.gov/ds/ 
si/ds/fspubschls.asp). That data file contains a vari- 
able that takes the following possible values: 

1 - Large city. A central city of Consolidated Met- 

ropolitan Statistical Area (CMSA), with the 
city having a population greater than or equal 
to 250,000. 

2 = Midsize city. A central city of a CMSA or Met- 

ropolitan Statistical Area (MSA), with the city 
having a population less than 250,000. 

3 - Urban fringes of large city. Any incorporated 

place. Census Designated Place, or non-place 
territory within the CMSA or MSA of a large 
city and defined as urban by the Census 
Bureau. 

4 = Urban fringes of midsize city. Any incor- 

porated place. Census Designated Place, or 
non-place territory within the CMSA or MSA 
of a mid-size city and defined as urban by the 
Census Bureau. 



5 = Large town. An incorporated place or Census 

Designated Place with a population greater 
than or equal to 25,000 and located outside a 
CMSA or MSA. 

6 - Small town. An incorporated place or Census 

Designated Place with a population less than 
25,000 and greater than 2,500 and located 
outside a CMSA or MSA. 

7 = Rural, outside MSA. Any incorporated place. 

Census Designated Place, or non-place terri- 
tory designated as rural by the Census Bureau. 

8 - Rural, inside MSA. Any incorporated place. 

Census Designated Place, or non-place ter- 
ritory within the CMSA or MSA of a large 
or mid-size city and defined as rural by the 
Census Bureau. 

To generate the urban, suburban, and rural catego- 
ries, these categories are mapped as follows: 

• Urban: large city or mid-size city. 

• Suburban: urban fringes of large city, urban 
fringes of mid-size city, large town, or small 
town. 

• Rural: rural, outside MSA, or rural, inside 
MSA. 



assessment of data quality 

Some data quality issues arise when working 
with the datasets described above. One challenge 
when working with California education data is 
that the state-level datasets are not always uni- 
fied or consistent. For example, school identi- 
fiers (such as County-District-School codes) are 
sometimes not provided and have to be generated 
from county, district, and school codes, add- 
ing an additional step to the process of merging 
school-level data. 

A second issue is that, given the several sources 
of data available (for example, CBEDS, API files. 



APPENDIX 



33 



TABLE A1 

Sizes of school quartiles by school characteristic 





Number of 
middle schools 


Number of 
high schools 


School quartiles by percentage of poverty (low-income students) 


Quartile 1 (lowest) 


306 


268 


Quartile 2 


305 


268 


Quartile 3 


306 


268 


Quartile 4 (highest) 


305 


268 


School quartiles by percentage of English language learners 


Quartile 1 (lowest) 


318 


287 


Quartile 2 


329 


270 


Quartile 3 


289 


288 


Quartile 4 (highest) 


312 


268 


School quartiles by percentage of Hispanic students 


Quartile 1 (lowest) 


320 


280 


Quartile 2 


313 


291 


Quartile 3 


303 


269 


Quartile 4 (highest) 


312 


273 


School quartiles by percentage of African American students 


Quartile 1 (lowest) 


320 


332 


Quartile 2 


370 


285 


Quartile 3 


264 


224 


Quartile 4 (highest) 


294 


272 


Population density 


Urban 


493 


385 


Suburban 


616 


472 


Rural 


94 


150 


School type 


Traditional public 


1,199 


967 


Charter 


51 


158 



Note: The quartiles do not have exactly the same number of schools because several schools have exactly the same percentage of English language learners, Hispanic students, 
and so on, making it impossible to define quartiles across schools with exactly the same value. Note that for each group the official percentages were used as provided in the 
2005 Academic Performance Index database, which reports percentages with two decimals. Therefore, the cut point between quartiles has only that level of precision. In addi- 
tion, the number of total observations for each analysis is different depending on the variables used, such as concentration of students eligible for free or reduced price lunch, 
English language learners, and so on, due to missing data in the CBEDS or API databases. 

Source: Authors' analysis based on data for 2005/06 from the California Basic Education Data System and other databases described in this appendix. 



CALWORKS, and STAR research files), it is pos- 
sible to have more than one value for the same 
variable. For instance, two different datasets of 
the PAIF from CBEDS provide information on 
whether a class satisfies UC and CSU require- 
ments (see the variable “UC_CSU_REQ” at 
www.cde.ca.gov/ds/ss/cb/fsasgncode.asp and the 



variable “UC_CSU” at www.cde.ca.gov/ds/ss/cb/ 
fsassignOO.asp). 

Finally, the typical problems of almost any data 
analysis (missing data and changing definitions 
of variables over time) are also present in this 
analysis. 





34 



THE DISTRIBUTION OF TEACHING AND LEARNING RESOURCES IN CALIFORNIA'S MIDDLE AND HIGH SCHOOLS 



REFERENCES 

AEL. (2003, December). Recruiting and retaining high- 
quality teachers in rural areas. Charleston, WV: Author. 
Retrieved January 13, 2007, from http://www.edvantia. 
org/publications/pdf/PB_Rural_Teachers.pdf 

Allgood, W., & Rice, J. K. (2002). The adequacy of urban 
education: Focusing on teacher quality. In C. F. Roellke 
& J. K. Rice (Eds.), Fiscal policy issues in urban educa- 
tion (pp. 155-180). Greenwich, CT: Information Age 
Publishing, Inc. 

Betts, J. R., Rueben, K., 8c Danenberg, A. (2000). Equal 
resources, equal outcomes? The distribution of school 
resources and student achievement in California. San 
Francisco, CA: Public Policy Institute of California. 
Retrieved January 13, 2007, from http://www.ppic.org/ 
content/pubs/report/R_200JBR.pdf 

Center for Education Reform. (2006, September). National 
charter school data: New school estimates 2006-07. 
Washington, DC: Author. Retrieved January 13, 2007, 
from http://www.edreform.com/index.cfm?fuseAction= 
stateStatChart8cpsectionid=158i:cSectionID=44 

Darling-Hammond, F. (1987). Teacher quality and equality. 
In J. Goodlad 8c P. Keating (Eds.), Access to knowledge: 
An agenda for our nation’s schools. New York: College 
Entrance Examination Board. 

The Education Trust-West. (2005, March). California’s 
hidden teacher spending gap: How state and district 
budgeting practices shortchange poor and minority stu- 
dents and their schools. Oakland, CA: Author. Retrieved 
February 10, 2007, from http://www.hiddengap.org/ 
resources/report031 105.pdf 

Education Week. (2006, June 22). Diplomas count: An es- 
sential guide to graduation policy and rates. Bethesda, 
MD: Editorial Projects in Education, Inc. Retrieved 
January 13, 2007, from http://www.edweek.org/ew/ 
toc/2006/06/22/index.html 

Fideler, E. F., Foster, E. D., & Schwartz, S. (2000). The urban 
teacher challenge: Teacher demand and supply in the 
great city schools. The Urban Teacher Collaborative. 



Retrieved January 13, 2007, from http://www.cgcs.org/ 
pdfs/utc.pdf 

Finn, J., & Achilles, C. M. (1999). Tennessee’s class size 
study: Findings, implications, and misconceptions. 
Educational Evaluation and Policy Analysis, 21, 97-109. 

Gandara, P., Rumberger, R., Maxwell-Jolly, J., & Calla- 
han, R. (2003). English learners in California schools: 
Unequal resources, unequal outcomes. Education Policy 
Analysis, 11(36). 

Goldhaber, D., & Brewer, D. (1996). Evaluating the effect of 
teacher degree level on educational performance. In W. 
Fowler (Ed.), Developments in school finance, 1996 (pp. 
197-210). Washington, DC: US. Department of Education, 
National Center for Education Statistics (ED 409 634). 

Governor’s Committee on Education Excellence. (2006, 
August). Getting down to facts: Studies and descrip- 
tions. Retrieved January 13, 2007, from http://www. 
everychildprepared.org/docs/studies_and_descriptions. 
pdf 

Greenwald, R., Hedges, L., & Laine, R. (1996, fall). The ef- 
fect of school resources on student achievement. Review 
of Educational Research, 66(3), 361-396. 

Guha, R., Campbell, A., Humphrey, D., Shields, P, Tiffany- 
Morales, J., & Weschler, M. (2006). California’s teaching 
force 2006: Key issues and trends. Santa Cruz, CA: The 
Center for Teaching and Learning. Retrieved January 
13, 2007, from http://www.cftl.org/documents/2006/ 
TCF2006FINAL.pdf 

Hall, D., & Kennedy, S. (2006, March). Primary progress, 
secondary challenge. Washington, DC: The Education 
Trust. Retrieved January 13, 2007, from http://www2. 
edtrust.org/NR/rdonlyres/15B22876-20C8-47B8-9AF4- 
FAB148A225AC/0/PPSCreport.pdf 

Hanushek, E. (1986). The economics of schooling: Produc- 
tion and efficiency in public schools. Journal of Eco- 
nomic Literature, 24(3), 1141-78. 

Ingersoll, R. (1996). Out-of-field teaching and educational 
equity. Washington, DC: U.S. Department of Education, 



REFERENCES 



35 



Office of Educational Research and Improvement 
(NCES 96-040). 

Kane, T. J., Rockoffi J. E., & Staiger, D. 0. (2007, winter). 
Photo finish: Certification doesn’t guarantee a win- 
ner. Education Next. Retrieved January 13, 2007, from 
www.tntp.org/newsandpress/newsitems/documents/ 
ednext_20071_60.pdf 

Milesi, C., & Gamoran, A. (2006). Effects of class size and 
instruction on kindergarten achievement. Educational 
Evaluation and Policy Analysis, 28{A), 287-313. 

National Center for Education Statistics. (2000, December). 
Monitoring quality: An indicators report. Washington, 
DC: Author. 

Nye, B., 8c Hedges, L. V. (1999). The long-term effects of 
small classes: A five-year follow-up of the Tennessee 
class size experiment. Educational Evaluation and 
Policy Analysis, 21{2), 127-142. 

Oakes, J., Rogers, J., Silver, D., Valladares, S., Terriquez, 

V., McDonough, R, Renee, M., 8cLipton, M. (2006). 



Removing the roadblocks: Pair college opportunities for 
all California students. Los Angeles, CA: University 
of California All Campus Consortium on Research 
and Diversity and University of California, Los An- 
geles, CA, Institute for Democracy, Education and 
Access. Retrieved February 10, 2007, from http:// 
www.idea.gseis.ucla.edu/publications/roadblocks/pdf/ 
ExecutiveSummary.pdf 

Oakes, J., Ormseth, T., Bell, R. M., 8c Camp, P. (1990). 
Multiplying inequalities: The effects of race, social class, 
and tracking on opportunities to learn math and science. 
Santa Monica, CA: RAND. 

Peske, H. G., 8c Haycock, K. (2006, June). Teaching inequal- 
ity: How poor and minority students are shortchanged 
on teacher quality. Washington, DC: The Education 
Trust. Retrieved January 13, 2007, from http://www2. 
edtrust.org/NR/rdonlyres/010DBD9F-CED8-4D2B-9- 
E0D-91B446746ED3/0/TQReportJune2006.pdf 

Sanders, W. L., 8c Rivers, J. C. (1996). Cumulative and 
residual effects of teachers on future student academic 
achievement. Knoxville, TN: University of Tennessee.