Does High-Stakes Testing Increase Cultural Capital among Low-Income and Racial Minority Students ?

This article draws on research from Texas and Chicago to examine whether highstakes testing enables low-income and racial minority students to acquire cultural capital. While students’ performance on state or district tests rose after the implementation of high-stakes testing and accountability policies in Texas and Chicago in the 1990s, several studies indicate that these policies seemed to have had deleterious effects on curriculum, instruction, the percentage of students excluded from the tests, and student dropout rates. As a result, the policies seemed to have had mixed effects on students’ opportunities to acquire embodied and institutionalized cultural capital. These findings are consistent with the work of Shepard (2000), Darling-Hammond (2004a), and others who have written of the likely negative repercussions of high-stakes testing and accountability policies.

future research on high-stakes testing and accountability.Our analysis indicates that testing and accountability policies in Texas and Chicago provided some opportunities for low-income and racial minority students to acquire key forms of cultural capital.At the same time, these policies also seemed to limit opportunities for such students and exacerbate differences between them and white, middle-class students.
In the first and second sections of this paper, we draw on cultural capital theory to present our conceptual framework and analytical methods.The third section reviews several studies of state policies in Texas to examine their effects on opportunities for low-income and racial minority students to acquire embodied and institutionalized cultural capital.In the fourth section, using studies of accountability policies in Chicago, we explore how such policies affected opportunities for such students to gain these forms of cultural capital.Finally, the fifth section looks across the two cases to discuss important findings and consider implications for future research on high-stakes testing and accountability.

Embodied and Institutionalized Cultural Capital
The conceptual framework employed in this study draws on cultural capital theory (Bourdieu, 1973(Bourdieu, , 1986)).In modern societies, according to this theory, social institutions such as schools may appear to be unbiased, neutral entities, but they are, in fact, governed by rules of exchange that place value on the cultural norms or cultural capital of upper class and middle class people (Bourdieu, 1986).Cultural capital exists in three states: embodied, institutionalized, and objectified (Bourdieu, 1986;Olneck, 2000). 1 We focus here on the first two of these forms because they seem most relevant to understanding the effects of high-stakes testing on low-income and minority students.
Embodied cultural capital refers to behavioral styles, ways of speaking, cultural preferences, and understanding of valued cultural knowledge (Olneck, 2000).Unlike high school diplomas, university degrees, or titles, this form of cultural capital cannot be purchased and unlike property, it cannot be exchanged; instead, it is learned or adopted by individuals.Bourdieu argues that schools do not value all students' speaking and behavioral styles equally, but rather they place greater value on those of the upper and middle classes.However, unless low socio-economic status (SES) and minority students have opportunities to internalize dominant cultural norms, they may be disadvantaged by their schools with regard to school engagement and performance, college attendance, and employment opportunity.Indeed, researchers have shown that the lack of cultural capital among low-income and minority students can result in reduced access to school resources and academic and social supports from teachers (Lareau, 2002;Lareau & Horvat, 1999;Lee & Bowen, 2006).For this reason, Delpit (1995) contends that schools should explicitly teach low-SES and minority students to acquire cultural norms, behavioral styles and codes of power that are necessary for them to succeed in American society.
In the 1990s and more recently, several policy makers and researchers have argued that a common curriculum linked to high-stakes testing could help low-income and minority students acquire the intellectual abilities and dispositions required in the 21st century societies.For example, O'Day and Smith (1993) posited that by making high quality knowledge and instruction available for every student, a common curriculum and a common set of expectations would decrease inequity in education by improving the performance of students from low-income and minority families.Also, Ravitch (1995) and Hirsch (1996) stressed that standards-based reform would enable American schools to accomplish what they had never done before: educate all students well, regardless of social class and racial backgrounds.More recently, advocates of high-stakes testing have contended that NCLB and similar state policies are necessary to ensure that teachers and schools maintain high standards for low-SES and minority students and help them achieve at high levels (e.g., Paige, 2001;Grissmer et al., 2000).
At the same time, scholars and educators have raised concerns that high-stakes testing and accountability policies will lead teachers to narrow the curriculum and devote inordinate amounts of time to preparing students to take state standardized tests (Shepard, 2000;Thompson, 2001).Darling-Hammond argues that overemphasis on test scores will lead to "a narrower curriculum; to test-based instruction that ignores critical real world skills, especially for lower-income and lowerperforming students; and to less useful and engaging education" (2004a, p. 18).Also, there is a growing concern that NCLB would interfere with teachers' efforts to develop relevant curriculum for culturally and racially diverse students (Selwyn, 2007).If this is the case, the new accountability system based on test scores is not likely to help low-income and minority students to acquire embodied cultural capital that is valued by universities and employers.
A second concern has been that disparities in resources severely limit the capacity of schools and districts serving high percentages of low SES and racial minority students.Researchers have documented significant differences with regard to school facilities and teacher quality between districts and schools serving primarily middle-class families and those mostly serving lower-income and minority students (Arsen & Davis, 2006;Lankford, Loeb, & Wyckoff, 2002;Loeb, Darling-Hammond, & Luczak, 2004).Under such conditions, teachers may not have the qualifications or resources to help students acquire embodied cultural capital in the form of analytical, higher-order thinking, and problem-solving skills.
Institutionalized cultural capital refers to degrees, credentials, grades, and test scores that serve as social markers to indicate that holders have specific levels or types of knowledge and skills (Olneck, 2000).As Labaree (1997) argues, a primary reason that individuals invest money, time, and effort in schooling is to acquire qualifications that will enable them to advance to higher levels of education and attain desirable employment and social positions.It is widely believed that schools are meritocratic with academic success being based solely on ability; according to this belief, schools provide each student with an equal chance to acquire academic credentials by using fair and objective methods such as grades and test scores.However, Bourdieu (1973) contends that the notion that schools are meritocratic is false and serves to legitimize the perpetuation of social hierarchies.
From Bourdieu's perspective, students from the middle and upper classes are more likely to succeed in school because they already possess the types of embodied cultural capital that schools value.Consequently, such students are more likely to acquire higher academic credentials and professional or white-collar jobs.In contrast, it is more difficult for low SES or minority students to succeed in schools.From the start, they have less of the embodied cultural capital that is necessary to thrive in schools.Even though some of these students may succeed academically through extraordinary efforts, the majority of them are more likely to fail or underperform in schools.As a result, in contrast to the ideology of equal chances, Bourdieu argues that schools contribute to reproducing existing social hierarchies.
Therefore, a key condition for high-stakes testing and accountability policies to succeed is whether they provide increased opportunities for low-income and minority students to acquire academic qualifications.This is a critical issue given the substantial number of U.S. students who drop out of high school and the significant gap between white students and African American and Hispanic students with regard to dropout rates.Mishel and Roy (2006) estimated that the overall high school graduation rate in the U.S. was between 80 and 83 percent in the 1990s and early 2000s.Further, they estimated that the graduation rate for African Americans ranged from 69 to 75 percent during this time while the graduation rate for Hispanics ranged from 61 to 74 percent (Mishel & Roy, 2006).
To the extent that dropout rates among low-income and racial minority students remain the same or increase under high-stakes testing policies, it seems likely that such policies would not be helping such students acquire institutionalized cultural capital.On the other hand, even when highstakes testing policies reduce the dropout rates among such groups, it would be important for low SES and minority students to also have opportunities to acquire embodied cultural capital to succeed in higher education and in their careers.

Methods
In selecting the major objects of analysis in this paper, we employed several criteria.First, we elected to focus on one state serving high percentages of racial minority students and one large urban school district to consider similarities and differences across these contexts with regard to the effects of high-stakes testing on cultural capital.Second, we were strongly interested in jurisdictions that had enacted high-stakes testing and accountability policies in the 1990s prior to the implementation of No Child Left Behind.The first two criteria led us to focus on Texas and Chicago in this article.
Texas was known as a leading state in this area in the 1990s and it served as a model for NCLB (Haney, 2000;McNeil, 2005).The state adopted the Texas Assessment of Academic Skills (TAAS) in the early 1990s, requiring students to take high-stakes tests in grades 4, 8, and 10.Based primarily on the percentage of students passing each of the TAAS tests, schools were rated as exemplary, recognized, acceptable or unacceptable.Schools were eligible for cash awards for high ratings, whereas underperforming ones were subject to sanctions, including possible closure. 2TAAS was a high-stakes test not only for schools but also for students.For example, if students did not pass all three portions of the exit level version of TAAS, they could not graduate from high school, regardless of their course grades.Only the scores on the state test counted as evidence in deciding whether a student was promoted or retained, and whether they earned a high school diploma or not (Haney, 2000;McNeil, 2005).
In the spring of 1995, the Chicago Public Schools (CPS) declared an end to social promotion, meaning that every student had to meet requirements to advance to the next grade.In 1996, a new accountability program took effect based on students' test scores on high-stakes tests.Under this new program, 3rd, 6th, and 8th graders had to meet minimum test-score standards in reading and mathematics on the Iowa Test of Basic Skills (ITBS) to move to the next grade.The cut scores were set to correspond to scoring roughly at the 20th percentile on national norms (Roderick, were required to participate in a summer school program and to retake the tests in August.Those who failed again at that time were to be retained in their grade. With regard to student demographics, in 1997-98, shortly after the implementation of TAAS, 45 percent of the 3.9 million PreK-12 students in Texas were white, 38 percent were Hispanic, and 14 percent were African American (Texas Education Agency, 1998).In 2005-06, of the 4.5 million PreK-12 students in Texas, 45 percent were Hispanic, 37 percent were white, and 15 percent were African American (Texas Education Agency, 2006).In 1995-96, when CPS declared its new promotional policy, 54 percent of its 407,000 students were African American, 31 percent were Hispanic, and 11 percent were white (Catalyst Chicago, n.d.).By 2005-06, CPS was serving about 420,000 students, 85 percent of whom were from low-income families, 49 percent of whom were African American, and 38 percent of whom were Hispanic (Chicago Public Schools, n.d.).
Third, each study featured in this article included measures of instruction and student learning (which are closely related to embodied cultural capital) or measures of students' test scores, high school graduation rates, or post-secondary outcomes (which are all forms of institutionalized cultural capital).
For both jurisdictions (Texas and Chicago), we drew on the theoretical framework described above to analyze the ways in which and the extent to which their high-stakes testing and accountability policies seemed to affect opportunities for low-income and racial minority students to acquire embodied and institutionalized cultural capital.Further, we considered whether these policies seemed to reduce, increase, or have no effect on differences in cultural capital between low SES, minority students and white, middle-class students.

Apparent Effects of Texas' Accountability System on Cultural Capital
High-stakes testing and accountability policy in Texas was the model for the federal No Child Left Behind legislation.Before his election in 2000, President Bush was the governor of Texas, and his first U.S. Secretary of Education, Rod Paige, had served as the superintendent of the Houston Independent School District.The close connection between NCLB and Texas continued with U.S. Secretary of Education Margaret Spellings who had worked on an education reform commission under Texas Governor William Clements and as associate executive director for the Texas Association of School Boards.As Grissmer andFlanagan (1998, 2001) argued, African American and Hispanic students made large gains on the Texas Assessment of Academic Skills (TAAS) and the National Assessment of Educational Progress (NAEP) in the 1990s, meaning that the Texas model for accountability seemed to contribute to reducing the gap among students who had different levels of cultural capital.However, other studies revealed that this accountability system seemed to have a mixed impact, including some harmful effects, on low-income and racial minority students (Haney, 2000;McNeil, 2005;McNeil & Valenzuela, 2001).
After Texas adopted its new accountability system in the early 1990s, test scores on the TAAS improved for all three of the major racial groups in the state: Hispanics, African Americans, and whites.For instance, Table 1 shows the percentage of grade 10 students meeting the TAAS minimum expectations between 1994 and 2002.The table demonstrates a notable increase in the percentages of 10th graders meeting the state expectations in all subject areas.In particular, there were substantial percentage increases among low-income and racial minority students.For example, in 1994, only 28 percent of African American students passed the 10th-grade TAAS exams; in 2002, 78 percent passed these exams.As a result of rapid gains among minority students, racial gaps in performance reportedly narrowed.In 1994, among all students in grades 3-8 and 10, only 31 percent of African American students passed the TAAS as compared to 66 percent of whites.By 1999, the gap between the two groups had been reduced to 24 percentage points and by 2002, it had shrunk to 15 percentage points (McNeil, 2005).A similar trend was found between Hispanic students and whites as well, which appears to support the notion that the new policy in Texas not only improved students' academic achievement but also reduced the gaps in achievement among students who may have entered school with different levels of cultural capital.
In contrast to these findings, other researchers have argued that these gains on TAAS and NAEP may not have represented the academic capacities of all Texas children (Gordon & Reese, 1997;Haney, 2000;McNeil, 2005;McNeil & Valenzuela, 2001).In particular, these scholars have shown ways in which the accountability system in Texas had strongly adverse impacts on lowincome and minority students.Based on survey and interview data from more than 100 teachers, Gordon and Reese (1997) reported widespread harmful effects of TAAS on low-income and minority students.The researchers contended that scores on TAAS were viewed by many of the teachers in their study as an objective criterion that confirmed at-risk students' failing performance; further, the results of the TAAS led many teachers to withdraw their support for such students when, in fact, these were the very students who needed additional support.
In a study that employed interview and observation data, McNeil and Valenzuela (2001) contended that TAAS had significantly changed the substance of schooling; i.e., what was taught and how students learned.They argued that during reading, writing, and mathematics instruction, teachers often placed greater emphasis on test preparation activities than implementing a rigorous, intellectually demanding curriculum.More recently, based on qualitative data collected from an urban elementary school in Texas, Booher-Jennings (2005) argued that teachers divided students into three groups-"safe cases," "suitable cases for treatment," and "hopeless cases"-with school resources distributed unequally to these groups.In her study, teachers targeted "suitable cases" who were on the threshold of passing the state tests and thus were likely to increase the school's aggregate passing rate, while withdrawing attention and resources from "hopeless" students.To the extent that this occurred, low SES and racial minority students lost opportunities to acquire embodied cultural capital in the form of higher-order thinking, analytical writing, and problemsolving skills.
In our analysis, we also focused on studies that examined the effects of high-stakes testing and accountability policy in Texas on opportunities for low SES and racial minority students to acquire institutionalized cultural capital.Despite its negative effects on these students' access to embodied cultural capital, the state's policy might still have had positive results if it led to improved test scores on the National Assessment of Educational Progress (NAEP) and lower dropout rates among these students.With regard to these two measures, the state's accountability policy seemed to have mixed results, including some negative consequences.Grissmer andFlanagan (1998, 2001) provided data showing that, along with North Carolina, Texas made the largest gains on the NAEP between 1990 and 1997.Further, even though both African American and white elementary students' scores increased, it appeared that African American students made larger gains than white students in reading and mathematics.The authors insisted that this progress was not due to traditional explanations of improved student outcomes such as per-pupil spending, teacher/pupil ratios, and percentages of teachers with advanced degrees or more years of experience.This was because Texas and North Carolina ranked at or below the national averages on these measures during the years of the study.Instead, Grissmer and Flanagan argued that the key reform policies associated with the NAEP gains in these two states were statewide academic standards by grade, holding all students accountable to the same standards, state-wide assessments closely linked to standards, and accountability systems based on student test scores.In short, they insisted that the rapid academic growth among Texas and North Carolina students was due to tightened accountability systems. 3ther researchers have raised concerns regarding Texas' apparent improvements on NAEP in the 1990s.First, in pointing out that Grissmer and Flanagan used NAEP reading data only for the 1992 and 1994 administrations of the exam, Treisman and Fuller (2001) included data from NAEP in 1992NAEP in , 1994NAEP in , and 1998 and argued that the gains among African American and Hispanic students between 1994 and 1998 were less significant.They also argued that the academic gains among Texas students were not just a product of the heightened accountability system but also of wider efforts such as equalizing school funding that Grissmer and Flanagan underestimated.Second, McNeil and Valenzuela (2001) pointed out that "the attention has been more on the rate of improvement on NAEP rather than actual improvement" (p.130).They argued that, even after the highly touted gains in mathematics and reading, the state's performance was still at or below average of national scores, registering lower than 21 of the 40 states participating in NAEP (McNeil & Valenzuela, 2001).
Further, Haney (2000) and McNeil (2005) argued that the state's NAEP gains seemed to be due to excluding students not expected to do well on the tests.In administering NAEP, school districts, schools and students were to be randomly sampled among participating states, but school personnel could remove students who they had classified as Limited English Proficient (LEP) or who had Individualized Education Plans (IEPs) as part of special education programs.McNeil (2005) argued that Texas had excluded higher percentages of students from taking the NAEP tests than most other states.From 1992 to 1996, exclusion rates in Texas increased from 8 percent to 11 percent at grade 4, and from 7 percent to 8 percent at grade 8.In contrast, the national exclusion rates decreased between 1992 and 1996 from 8 percent to 6 percent at grade 4 and from 7 percent to 5 percent at grade 8. 4 In sum, while Grissmer andFlanagan (1998, 2001) argued that high-stakes testing led to Texas students' rapid gains on NAEP, it appears that the state's substantial exclusion of students may also have affected its performance on this national assessment.
In addition, we need to consider that the high school graduation rate in Texas is fairly low; in a 2001 report, it was 67 percent, ranked 40th in the nationwide order (Greene, 2002).Table 2 shows high school graduates in 2000-2001 as a percentage of average enrollments in grades seven to nine in the four largest school districts in Texas.The data in Table 2 indicates that roughly half of the students who were in 9th grade in 1997-98 in these four districts did not graduate from high schools in these districts in 2000-2001.The racial gap in high school graduation rates also increased after the implementation of TAAS.Between 1992 and 1998, the ratio of high school graduates to grade 9 students three years earlier had been at or below 0.50 for African American and Hispanic students, while it had been about 0.70 for white students (Clarke et. al., 2000).According to Haney (2000), from 1978 through 1989, the average gap between the ratio for white students and the ratio for African American and Hispanic students had been 0.146.However, the average gap between the ratio for whites and that for non-whites students grew to 0.215 after the TAAS exit test requirement was fully implemented in 1992-1993.This indicates that the TAAS exit test had caused a 50 percent increase in the gap between the ratio for white students and the ratio for non-white students.McNeil argued that the Texas accountability system itself "creat(ed) incentives for principals to 'lose' their low-performing students, more frequently their Latino, African American, LEP, and immigrant children, to make sure the schools' scores are high" (2005, p. 74).She also contended that a widespread waiver system in Texas schools was the primary cause of this large loss of students; to improve the performance of their schools on the 10th-grade exams (i.e., high school exit exams), many administrators, teachers, and counselors held African American and Hispanic students back in 9th grade (Haney, 2000;McNeil, 2005).
In sum, Texas' accountability system had mixed effects on opportunities for low SES and racial minority students to acquire cultural capital.While some researchers reported that students' scores on TAAS increased in the 1990s and racial gaps in achievement decreased, other studies raised concerns about the effects of high stakes testing on classroom instruction and high school graduation rates.In particular, the accountability system seemed to dramatically alter the substance of schooling and to lead many districts and schools to exclude students from taking the state tests.Further, the racial gap in high school graduation rates increased following the implementation of TAAS.As a result, while the state tests provided opportunities for some low-income and minority students to acquire cultural capital, they led many others to leave school without their diplomas-a key form of institutionalized cultural capital.

Apparent Effects of Chicago's Accountability System on Cultural Capital
The Chicago Public Schools (CPS) is the nation's third largest school district and it predominantly serves low-income and racial minority students.In the spring of 1995, CPS declared an end to social promotion, meaning that every student had to meet requirements to advance to the next grade.In 1996, a new accountability program took effect based on students' test scores on high-stakes tests.Under this new program, 3rd, 6th, and 8th graders had to meet minimum testscore standards in reading and mathematics on the Iowa Test of Basic Skills (ITBS) to move to the next grade.The cut scores were set to correspond to scoring roughly at the 20th percentile on national norms (Roderick, Jacob & Bryk, 2002).Students who did not meet the cutoff standard at the end of the school year were required to participate in summer school programs and to retake the tests in August.Those who failed again at that time were to be retained in their grade.
In the case of grade 8, those who did not pass the second time either were retained or moved into a transition center.Transition centers were new schools designed for 8th graders who failed to meet the promotion requirement but were too old-more than 15 years old-to remain in elementary schools. 6As a result of the new promotion policy, more than one third of grade 3, grade 6, and grade 8 students failed to meet the promotion cutoffs in the first two years (Roderick, et al., 1999).Roderick and Nagaoka (2005) reported that, among Chicago students in the gate grades (i.e., grades 3, 6, and 8), 7,000 to 10,000 students were retained each year between 1996 and the early-2000s.
Proponents of retention policies argued that establishing cutoff standards, making clear that achievement matters, and imposing negative consequences would lead students to work harder and teachers to pay attention to the needs of the lower-performing students.They also contended that if students have not mastered basic skills, they would be better served by repeating the same grade and gaining those skills than by struggling with more advanced materials (Roderick & Nagaoka, 2005).In other words, it was expected that making clear what needs to be taught for the students to pass the promotion gate would help them acquire embodied cultural capital valued in schools and society.Indeed, after the inception of the new accountability program, Chicago students' test scores rose to some extent and the proportion of students in the gate grades with test scores below the minimum standard for promotion fell significantly.For example, in grade 6, the percentage of students who failed to meet the promotional cutoff fell from 37 percent in 1995, the year before the policy took effect, to 14 percent in 1999 (Roderick & Nagaoka, 2005).Further, using achievement data from 3rd-, 6th-, and 8th-graders between 19903rd-, 6th-, and 8th-graders between and 20003rd-, 6th-, and 8th-graders between , Jacob (2002) ) found that students' scores on the ITBS reading and math tests consistently increased after the implementation of the new promotion policy.
As in Texas, though, the new retention policy in Chicago seemed to have a negative impact on opportunities for low SES and racial minority students to acquire embodied cultural capital.Through a further analysis of math scores, Jacob (2002) argued that the large ITBS gains were driven primarily by questions testing basic skills, such as computation and number concepts, which were easier to teach through the district's ITBS-specific curriculum.In contrast, students made very little improvement on questions requiring more complex skills such as estimation, data interpretation and problem solving.He also found that students' math achievement on the ITBS tests had little relationship with their performance on the state IGAP (Illinois Goals Assessment Program) tests, which placed more emphasis on critical thinking and problem-solving skills (Jacob, 2002).
Other studies also corroborate the contention that improved test scores do not mean that students have learned higher-order thinking and academic skills.Based on qualitative data from four Chicago elementary schools and interviews with CPS district administrators, Lipman (2004) reported that test preparation for students had replaced potentially rich educational experiences in many urban schools.In her words, "students spend hours taking mock tests, practicing filling in bubbles in scantron sheets, developing familiarity with the layout of the tests and the kinds of questions that are asked, and learning 'tricks' for eliminating incorrect answers" (Lipman, 2004, p. 79).As in Texas schools, the focus on reading and mathematics testing had led schools to place less emphasis on social studies, science, and other subjects that were not tested in the accountability system.In one school, for example, Lipman (2004) observed that teachers stopped teaching social studies and concentrated on reading and mathematics instruction from January through May.
Lipman also contended that while more affluent students in Chicago were engaged in intellectually challenging curriculum, low-income and minority students had to memorize fragmented facts and information and master simple test-taking techniques.She argued that this differential access to high-quality curriculum had significant consequences for social inequalities in an information-based economy.That is, white students who already possessed embodied cultural capital valued by schools were more likely to advance to higher education and attain professional, managerial and technical jobs.On the other hand, low SES and minority students in urban schools were more likely to take low-level, low-skill, and low-paid jobs in Chicago's growing service economy.As a result, differentiated access to rigorous curriculum seemed to contribute to reproducing the asymmetrical social structure.
In another study, Anagnostopoulos (2006) observed demoted 9th graders' classrooms in two Chicago high schools,7 providing evidence regarding the unequal distribution of opportunities to acquire embodied cultural capital.Rather than compelling teachers and students to remedy school failure academically, the CPS retention policy facilitated a type of moral boundary work that distinguished "deserving" students from those deemed "undeserving" (Anagnostopoulos, 2006).In particular, teachers and students did not regard test scores as an expression of academic abilities.Rather, they employed a moral judgment to draw a line between promoted and demoted students.That is, both teachers and students "described demoted students as 'lazy,' 'apathetic,' 'disruptive' and even 'criminal'" (Anagnostopoulos, 2006, p. 17).In other words, teachers and students did not think that demoted students had difficulties in learning and thus needed additional support.Instead, they believed that they did not work hard enough and thus did not deserve care and attention.
From the teachers' perspectives, the moral designation of demoted students justified placing them at the margins of the school's moral order and withdrawing instructional resources from them.For example, teachers spent a higher percentage of time on management in demoted classes than in promoted or mixed classes.On average, teachers in the demoted classes spent almost 20 percent of class time on management and discipline whereas they spent approximately 14 to 15 percent of allotted time on these activities in promoted and mixed classes (Anagnostopoulos, 2006).Further, students in demoted classes spent fully 25 percent of their time in class engaged in non-academic activities.Consequently, demoted students who had already been marginalized in school were further excluded from access to embodied cultural capital, which contradicted the promise made by advocates of high-stakes testing: that a challenging and rigorous curriculum would be provided to all students.
With regard to institutionalized cultural capital, researchers have shown that the promotional policy in Chicago had disproportional effects on low-income and racial minority students.When CPS placed 109 schools on probation-71 elementary and 38 high schools-for their low passing rates in 1996, the average poverty level of the 71 elementary schools was about 94 percent, meaning that probation schools were overwhelmingly African American and Hispanic schools (Lipman, 2004).In 1998, among 8th graders enrolled in transition centers, 71 percent were African American students, 25 percent were Hispanic students, and 3 percent were white students (PURE, 1999).8This higher rate of grade retention among African American students seemed likely to eventually result in higher dropout rates.Based on data from 1992 to 1998, Allensworth (2004) estimated that retained students' likelihood of dropping out increased by 8 percent by age 17 and by 13 percent by age 19.As a result, she argued that, whereas white and Hispanic students' dropout rates had declined in those years, African American students' dropout rates did not show any significant changes.
Other research studies indicate that the CPS policy provided little advantage to retained students.For example, comparing the achievement growth of students whose test scores fell just above and just below the test-score cutoff, 9 Roderick and Nagaoka (2005) contend that retention did not provide significant academic benefits to 3rd graders who were retained and had more negative effects on 6th graders who were retained.In their study, 3rd graders in the retained below-cutoff group experienced a slight boost in performance in the post-gate year.This effect, however, was small and short-lived; within two years, the achievement growth of the below-cutoff group was not statistically different than that of the above-cutoff group.Further, according to the authors, within two years of the gate grade, the possibility for retained 3rd graders to be placed in special education was almost three times higher than that of other low-achieving students (Roderick & Nagaoka, 2005).
In the case of 6th graders, the authors found that retention was associated with more negative growth in achievement one year after the gate grade, with that effect remaining two years later.Retained 6th graders were placed in special education at more than six times the rates of other low-achieving students.Roderick and Nagaoka (2005) argue that these findings were consistent with other research findings that retention had more harmful effects on matured students who were more sensitive to their social reputation.These findings suggest that the CPS retention policy had different effects on two groups of students-one retained and the other promoted-who in fact had similar academic capacities.Roderick and Nagaoka (2005) presume that this was because most retained students were concentrated in low resource schools that could not afford to provide additional supports for them.They also point out that CPS provided little guidance to teachers in diagnosing retained students' learning difficulties, devising effective instructional strategies, or providing learning materials.Rather, the "basic theory of action was that a second dose of the same material would be enough" (Roderick & Nagaoka, 2005, p. 332).
In sum, despite some gains in students' academic achievement, the promotional policy in Chicago may not have helped low-income and minority students acquire necessary cultural capital; instead, it seemed to narrow the curriculum and make it harder for those students to acquire higherorder thinking, writing, and problem-solving skills.Further, researchers found that retained students were disproportionately low-income and minority and that these students experienced few academic benefits and many social challenges.As a result, it seemed unlikely that being retained enabled them to acquire embodied and institutionalized cultural capital.

Conclusion
In recent years, researchers in the U.S. have drawn on Bourdieu's notion of cultural capital (1973,1986) to examine a number of educational issues.For example, Lareau has done pioneering work on how cultural capital influences home-school relations and parent-child interactions (Lareau, 2002;Lareau & Horvat, 1999) while Olneck (2000) has analyzed whether multicultural education curriculum and schooling practices are likely to redefine what constitutes cultural capital.This study contributes to this body of scholarship by considering the effects of high-stakes testing and accountability on minority and low-income students' access to cultural capital.In particular, we drew on research from Texas and Chicago to investigate the effects of such policies on students' opportunities to acquire embodied and institutionalized cultural capital.
In both contexts, there were indications that students' performance on state or district tests increased following the implementation of high-stakes testing and accountability policies.In Texas, the performance of all racial groups increased on both TAAS and NAEP during the 1990s.Similarly, the performance of students in the gate grades (i.e., grades 3, 6, and 8) improved after the enactment of the new accountability program in 1996.At the same time, research from Texas and Chicago revealed that these accountability policies seemed to have had deleterious effects on curriculum, instruction, the percentage of students excluded from the tests, and student dropout rates.These findings are consistent with the work of Shepard (2000), Darling-Hammond (2004a), and others who have written of the likely negative repercussions of high-stakes testing and accountability policies.
There are some limitations to our analysis and the studies included herein.First, most studies of the effects of accountability policies on curriculum and instruction included small samples of teachers and schools.In future studies on this topic, it would be advisable for researchers to include larger numbers of teachers and schools and to control for the possible effects of school and student characteristics.Second, these studies from Texas and Chicago focused on policies first implemented in the 1990s.It will be important for future research on high-stakes testing to consider the impact of NCLB itself in particular state and district contexts.Finally, data on dropout rates is often subject to multiple interpretations because of variability in the way such rates are measured and difficulty in obtaining reliable data.At the same time, to assess the effects of testing and accountability policies on institutionalized cultural capital, it will be necessary for researchers to have good data on high school completion rates and college attendance rates.
Despite these limitations, the research findings from Texas and Chicago seem to provide a number of guidelines as policy makers consider ways to revise NCLB as part of reauthorizing Title I in 2008 and 2009.First, there is a need to consider how high-stakes testing and accountability policies influence curriculum and instruction.In particular, research from Connecticut suggests that testing policies can be designed to promote cognitively advanced instruction while holding schools and districts accountable for performance (Darling-Hammond, 2004b;Youngs & Bell, 2007).Second, policy makers need to ensure that districts and schools do not respond to accountability policies by excluding students from taking tests.Finally, they need to continue exploring ways to make high school a meaningful, engaging, and cognitively demanding experience for all students (Sizer, 1984).While high-stakes tests provide important information about student performance, they will not, by themselves, lead to the types of reform at the high school level that would enable greater numbers of low-income and racial minority students to acquire embodied or institutionalized cultural capital.

Won-Pyo Hong
Hong is a doctoral candidate in the Department of Teacher Education, Michigan State University.His research interests center on how curriculum policy, accountability systems and standards-based reform have differential impacts on students from diverse socioeconomic and cultural backgrounds.Peter Youngs is an assistant professor in the Department of Teacher Education at Michigan State University.His research interests focus on policy and practice in the areas of teacher education, induction, and professional development.Recent publications have appeared in Teachers College Record, Educational Administration Quarterly, and Studying Teacher Education: The Report of the AERA Panel on Research and Teacher Education.

Table 1
Grade 10 Students Meeting TAAS Minimum Expectations in 1994 and 2002

Table 2
High School Graduation Rates in Major School Districts in Texas