I. Introduction
Some of the persistent issues on education in the United
States are regarding the teacher scarcities and teaching quality.
There is a seeming failure for school officials to attract
top candidates to teaching and retain qualified teachers. This may be due to
the surprisingly gender unification of the labor market. Educational sector
once having a female-filled labor pool is now facing competition against
profitable occupations thus the best and brightest are least expected to stay
in teaching. Such observation has been considered as one of the catalysts for
certain policy measures such as wage hikes, relaxed requirements, etc.
The emphasis on the importance of the quality of teaching on
student achievement is due to the lack of significance of other inputs into
schooling (per student spending and class size). Having perceived the weight of
attracting quality teachers, the federal government acted in 2001 through the
Title II of the Elementary and Secondary Education Act which is a $3.175
billion program for the purpose of improving teacher quality.
The assumption that gender desegregation of professions
caused the reduction of highly skilled female choosing to teach is almost a
common knowledge due to its repeated appearance in several studies. However the
degree to which occupational desegregation has in fact affected quality of
teaching is still uncertain.
II. Background
During the period 1962- 1992 the labor force participation
among women has extensively changed: number of women aged 25-34 roughly doubled
and a portion of which having a (four-year) college degrees tripled. In
contrast, over the same period, the fraction of college graduate men rose by
only 50 percent. Such developments may have been partially aided by key
legislative movements of the early 1960’s and 1970’s that radically changed the
occupational opportunity landscape for women.[1]
Over the period the labor force became substantially less
gender segregated. It was observed that
in the late 1960’s, in medicine and law, women were severely underrepresented,
however, in less than thirty years, for both occupations women are close to
being equally represented.[2]
Regardless the remarkable changes in the gender composition
of professions such as medicine and law, for the case of new teachers, it has
remained rather steady. Females still dominate the teaching occupations, about
70 percent of teachers aged 25-34. For female teachers, they comprise the
elementary faculty by about 82-85 percent, while only roughly 46-57 percent in
the secondary – however, it was observed that within the age range, the portion
of teachers has been consistently declining for elementary and increasing for
secondary.
Though teaching is still somewhat dominated by female, the
importance of the profession as a career path for female college alumni has
considerably declined – such drop can be credited to the massive increase in
college completion among women. The clear observation from such finding is that
contingent on working, most of those female graduates in the 1960’s went into
teaching while most of those competing college today do not.
However business desegregation and shift of females into
high skill occupations do not automatically imply a decline in the teaching
pool, if the pool of college graduates is growing. Between 1960 and 1996 the
total stock of teacher rose merely by 89 percent, with the increase mostly in
the period of 1960 to 1976 to provide for the enrollment boom. On the other
hand the share of female college degree holders tripled.
III. Prior
Literature on Teacher Quality
Weaver (1983) reported about 100-year history of efforts in
raising teacher standard in the United States. Most of the study on teacher
quality concentrated on:
1. Defining
quality and method in measuring it
2. Comparing
new entrant in teaching along a range of dimensions to their non-teacher
counterparts
3. Significance
(or lack thereof) of teacher attributes to student achievement
4. Determining
the way to enhance the quality of the teaching force without drastically
dropping the teacher supply or the diversity of the occupation.
Preferably, teacher quality would be gauged as a
multi-dimensional vector of all teacher attributes positively related with
educational process outputs. Since some of the characteristics such as
patience, creativity and so forth are unobservable or difficult to measure,
research was focused more on the measurable attributes such as degree of
attainment, certification status, teaching experience, and test scores. Other
studies make use of relative teacher salary as proxies for teacher quality.
For this paper, a measure of academic ability – scores from
a standardized test administered in high school – is used as measure of
“quality”. Though it may be considered
naïve to think that a single test score can capture all of the aspects that
comprise for an effective teacher, this study is able to show that scores
certainly seize “something” that is significant in explaining the academic
achievement of their pupils.
On the other hand, several studies have shown that the
teacher’s (and would-be teacher’s) academic ability in terms of their SAT or
ACT scores, fails in comparison to that of their peers (who do not major in
education) in the 1970’s and 1980’s. Though such comparison may be informative,
it can also be ambiguous as not all SAT/ACT go to college, and those who attend
may change majors or may not at all pursue teaching.
Vance and Schlechty (1982) were the first ones to use a
longitudinal study in comparing the teachers and non-teachers academic ability.[3]
The use of a longitudinal data actually follows pupils into the labor force and
helps avoid the problems from using SAT scores of future education majors.
Their study showed that teachers recognized in 1979 came disproportionately
from the bottom two quintiles of the SAT score spread.
A number of other cross-sectional researches have observed a
negative association between academic ability and the possibility of going into
teaching among college graduates. This suggests that during the 1970’s and
1980’s, the test scores of those students considering teaching fails in
comparison to those of their college peers.
In addition, with regards to the changes in the quality of
teacher over time:
1. Murnane,
et. al.(1991) and Bacolad (2001):
- The
percentage of graduates going into teaching decline 1967-1989, majority of the
decline was among those with high IQ scores.[4]
2. Lakdawalla
(2001):
- The
study interprets that the decline in the teacher’s education and human capital
value in relation to other workers is a decrease in the teacher quality due to
the substitution of the quantity for the quality of teachers.
3. Stoddard
(2001)
- The
author infers the same decline in the wage rank of teacher salaries as a drop
in teacher quality.
IV. Methodology and
Data
A. Methodology
Data from four longitudinal studies were combined to
understand the changing relationship between academic ability and entry into
teaching. In each of the four groups, those individuals having at least a high
school diploma is included in the sample. Although it is known that any alteration
in the ability composition of high school alumni can bias the results of the
paper, there is no means through the available data to distinguish the degree
of change in the composition, however it is believed that any bias brought by
such changes is smaller than the likely bias arising from limiting the analysis
to college graduates.
B. Data
Data comprises of four separate groups of high school
graduates from four longitudinal surveys: Wisconsin Longitudinal Study (WLS)
1957 class; Project Talent (Talent), 1960-1964 classes; National Longitudinal
Study of the High School Class of 1972 (NLS-72); and the sophomore group of
High School and Beyond (HSB) 1982 class.
These four studies include a detailed survey of the pupils
during their senior (or sophomore) year, and required them to partake in a
series of aptitude tests and all carried out several follow-up surveys after
high school. Overall, these surveys give four distinct representations of
relationship of academic ability with occupational choice among high school
graduates over four decades.
To permit assessment across surveys and avoid life-cycle
impacts on profession choice, the follow-up survey carried out when most
respondents were roughly 26 years old was chosen. Students’ combined raw scores
on math and verbal parts of the exam administered in high school were used as
the measure of cognitive or academic ability. After performing logistic
regression, it was found that the coefficient on test score is positive and
statistically significant which implies that the qualities measured by the
scores are strongly related with entering into a high-cognitive skill
occupation such as medicine. Also the marginal impact of test score on entry
into medicine among men stays approximately steady across four surveys.
On the other side, Polachek (1981) and Flyer and Rosen
(1997) implied that women expected to spend more time not in the work force may
choose occupations such as teaching due to their flexibility. Taking this into
consideration, comparing teachers to non-teachers in a sample of working women
may cause biased outputs.
Note that the data has no accurate distinction between the
classes of teachers (if elementary, or secondary, work or have worked in public
or private school districts).
Also, the results from the study suggest that though the
academic ability of the average new female teacher has stayed relatively
unchanged with respect to the mean high school graduate, it is still rather
unlikely that teachers will come from the top decile of their high school class.
V. Results – New
Entrants into Teaching
A. Female
Entrants
Finding showed that there was substantially less
concentration in the professions held by female in 1992 (in comparison to
1964). The drop may be due to the noticeable decline in the possibility of
teaching among women at the top end of the test score spread.
The results showed that for the sample graduates (for four
groups) having at least a high school degree here is a positive and
statistically significant association between test scores and going into
teaching. Another striking observation is the noticeable waning of the
relationship since 1964.
Through further analysis it was detected that there is a
nonlinear relationship between test score and entry into teaching, also, the
somewhat larger decline in the expected likelihood of becoming a teacher among
those of the high skill group. In 1992, roughly 3.7 percent of top females at
the sample opt to teach, a value significantly lower than the 20 percent
reported in 1964. Top scoring women in 1992 were more inclined to work as
computer specialist, accountants, or managers by 5.9 percent, 6.0 percent and
15.1 percent, respectively; in addition their possibility of becoming lawyers
and judges (roughly 3.2 percent) is almost the same as becoming a teacher.
B. Male
Entrants
An interesting by-product of the gender desegregation of
professions and the shift of women into high-cognitive skill occupations is
probably the replacement of talented men into teaching. Some men may have gone
into teaching due to finding themselves in competition with skilled women for
position and the diminishing of the male’s virtual monopoly on certain
occupations.
The results showed that for 1964-1992 the mean male
teachers’ academic ability increased. In addition, like that of women, the
possibility that a male high school graduate went into teaching declined;
however the drop in the probability for those in the top decile is less
striking.[5]
VI. Conclusion
There has been little empirical proof on how certain attributes
such as teachers’ academic abilities have altered over time.
Some of the results:
1. At
least among the four groups of high school graduates, there was a slight fall
in the academic ability of the average new female teacher
2. The
possibility of going into teaching by age 25-26 has declined for all ability
levels of high school graduates – and dropped considerably more for those
female belonging at the top distribution. An opposite trend was observed among
men, however it must be remembered that male sample size is smaller.
Though such results are indicative of trends in teacher
quality among new female teachers, and that the estimates give preliminary
measures of the degree to which females with high ability are less probable to
become teachers, there are still some flaws innate in the data:
1. Though
the four surveys all have advantage of being longitudinal in nature, such data
only gives snapshots of just four specific moments in time.
2. The
exam provided to students differed for each survey.
3. Test
scores are merely just one measure of teacher “quality”.
Applying the paper’s results to a broader population of new
teachers in the United States, with respect to having female teachers, a
student in 1992 could expect to have a teacher who is approximately only slightly
below academic ability than a given student in 1964. Moreover, the possibility
that a student in low income or mainly black school would have teachers of high
academic ability is likely to be low.
Source:
Sean P. Corcoran, William
N. Evans and Robert S. Schwab, “Changing Labor Market Opportunities for Women
and the Quality of Teachers 1957-1992”, NBER
Working Paper 9180, September 2002.
[1] Some examples of the legislative
movements are Equal pay Act of 1963 and Title VII of the Civil Rights Act of
1964, which explicitly punish employer sexual discrimination (Lloyd and Niemi, 1979).
[2] A value below one (for the
index) states that women are underrepresented. In the 1960’s the indexes were
roughly .2 and 0 for medicine and law, respectively. However after less than
thirty years it rose to .79 and .94, correspondingly.
[3] National Longitudinal Study of
the High School class of 1972
[4] National Longitudinal Studies of
Young Men, Women and Youth
0 comments:
Post a Comment