Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 Feb 21.
Published in final edited form as: Ann Am Acad Pol Soc Sci. 2010 Jan 4;627(1):82–105. doi: 10.1177/0002716209348741

Policy Transparency and College Enrollment: Did the Texas Top 10% Law Broaden Access to the Public Flagships?

Mark C Long 1, Victor B Saenz 2, Marta Tienda 3
PMCID: PMC3578658  NIHMSID: NIHMS410775  PMID: 23436938

Abstract

By guaranteeing college admission to all students who graduate in the top 10% of their high school class, H.B. 588 replaced an opaque de facto practice of admitting nearly all top 10% graduates with a transparent de jure policy that required public institutions to admit all applicants eligible for the guarantee. The transparency of the new admission regime sent a clear message to students attending high schools that previously sent few students to the University of Texas at Austin and Texas A&M University. Using 18 years of administrative data to examine sending patterns, we find a sizeable decrease in the concentration of flagship enrollees originating from select feeder schools and growing shares of enrollees originating from high schools located in rural areas, small towns, and midsize cities, as well as schools with concentrations of poor and minority students. We also find substantial year-to-year persistence in sending behavior once a campus becomes a sending school, and this persistence increased after the top-10% policy was implemented.

1. Introduction

In justifying H.B.588, the bill’s chief architect, the late Irma Rangel, emphasized that public institutions should be available to all Texas residents, irrespective of socioeconomic circumstances, ethnic group membership, or geographic location (Giovanola, 2005). Yet, with few exceptions (Montejano, 2001; Saenz, 2007; UT-OAR, 2008), attention to the socioeconomic and geographic consequences of the law has been limited. Because the top 10% law was implemented in response to the judicial ban on affirmative action, most evaluations of its impact have focused on its effectiveness in restoring campus ethno-racial diversity, particularly at the public flagships (see, for example, Horn and Flores, 2003; Kain and O’Brien, 2003; Niu, et al., 2006; Long and Tienda, 2008a; 2008b; Niu and Tienda, 2008; Tienda and Niu, 2006a).

Both because the law is allegedly race neutral (but see Forest, 2002; Tienda and Niu, 2006a) and because it grants college access by rewarding academic achievement (class rank), the top 10% admission regime was initially applauded as a viable alternative to affirmative action; however bi-partisan support for the law has eroded since its enactment (Hughes and Tresaugue, 2007; Monastersky, 2007; Tienda and Sullivan, 2009). That students eligible for automatic admission were qualified on a school-specific basis is a key provision behind growing opposition to the law. Given large economic inequities and racial segregation of Texas public schools, this provision is the linchpin for broadening geographic, socioeconomic and ethno-racial diversity as Rangel envisioned. Perceptions that high-achieving students from low-performing schools gain access to the public flagships at the expense of lower ranked students who graduate from high-performing schools foment resentment from parents who presume their students are being crowded out by less meritorious applicants (Niu and Tienda, 2008; Tienda and Niu, 2006b).1 As growing numbers of students from affluent suburban districts that have historically been major feeders to the public flagships are denied admission, calls to repeal the law during the 81st legislative session have resurfaced (University of Texas, 2008; Root, 2009; Jaschik 2009).

Most academic research about the top 10% law shows only modest increases in minority representation at the public flagships (Kain and O’Brien, 2003; Forest, 2002; Long and Tienda, 2008a), but there is emergent evidence that the policy changed students’ application behavior in ways that depend both on their class rank and type of high school attended (Long and Tienda, 2008b; Koffman and Tienda, 2008). The growth in applications to the University of Texas at Austin (UT) from students who graduated in the top-10% of their high school classes increased the shares of the freshman class automatically admitted from 41 to 81 percent between 1997 and 2008 (Schevitz, 2008; University of Texas, 2008). Concomitantly, high school sending patterns also appear to have changed. Between 1992 and 2002, the number of high schools that sent one or more applicants to the University of Texas at Austin (UT) rose from 678 to 798; at Texas A&M University (TAMU), the comparable increase was from 819 to 925 schools.2

These stylized facts provide some evidence that the uniform admission policy broadened geographic access to the two public flagships, a conclusion also consistent with Montejano’s (2001) claim that the initial impact of the new admission regime was geographic. Three years after the top 10% policy was in place, Montejano observed an emergent sending pattern involving new high schools that previously had sent few if any students to the University of Texas at Austin (UT). These students hailed from districts that largely served urban minority communities, such as inner city Houston, and poor white communities in northeast and west Texas. Montejano interpreted these incipient trends as evidence that the new admission regime supported one of Rangel’s goals of reaching a broader geographic spectrum of the State’s residents. More recent data indicate that the expansion of high school campuses that feed UT-Austin’s enrollment has continued, even while the actual number of enrolled students has remained relatively stable.3 In 2007 UT admitted students from over 900 different high schools across the state, up from 674 high schools in 1996, the year before the uniform admission regime was implemented (Saenz, 2007).4

Although these changes seem to imply broadened access, it is also conceivable that these increases simply reflect growth in the total number of high schools in Texas. The State’s school-age population has been growing much faster than the national average (WICHE, 2008), prompting the opening of new high schools. Therefore, claims about expanded access as a result of the top 10% policy must consider changes in the shares of high schools sending students to the public flagships.

Sending a single enrollee is a weak criterion for measuring a high school’s “access” to the flagships. Whether growth in the number of high schools represented among UT and TAMU enrollees truly represents broadened access to “all Texas residents,” as Rangel envisioned, also depends on the school’s application and enrollment rates. Traditionally, the vast majority of high schools sent none or a handful of enrollees to the flagships, while a few high schools sent very large numbers of students. Tienda and Niu (2006b) distinguish between “sending” schools (those that send one or more applicants) and “feeder” schools (the top-twenty high schools based on the number of students admitted to UT and to TAMU as of 2000). Because of substantial overlap in the college destinations of their graduates, feeder schools represented only 28 unique secondary campuses out of over 1500 statewide. In 2000 the 28 feeder schools accounted for about 15 and 23 percent of freshman enrollment at TAMU and UT, respectively. Saenz (2007) reports a similar concentration of feeding patterns to UT. He shows that half of UT’s 1996 freshman class came from 59 high schools, but by 2006, half of UT’s enrollment came from 104 high schools. The top feeder schools use strategic resources, such as well-placed alumni networks, strong counseling offices, and well educated parents, to maintain close relationships with the State’s flagship institutions.

These shifts in sending patterns put into context the current controversy about the fairness of the uniform admission law. Specifically, legislators from sparsely settled rural districts allege that the provisions of the law guaranteeing admission to rank-eligible graduates “reserves” slots for them that would not be available otherwise; detractors emphasize that students from rural high schools represent less than two percent of total enrollment (see Hughes and Tresaugue, 2007; Monastersky, 2007). These legislative debates about whether the uniform admission law broadened geographic diversity at the public flagships by changing the sending patterns of applicant and enrollment pools show no signs of abating. UT, in particular, has been at the center of most controversy about the law because it has become saturated with top 10% admits, leaving little flexibility for administrators to shape the composition of freshman classes (Paredes, 2006; Tienda and Sullivan, 2009; University of Texas, 2008).

Evaluating competing claims about increased geographic access based on changes in high school sending patterns is crucial for informed policy decisions (University of Texas, 2008; Root 2009; Jaschik, 2009).5 Accordingly, we use administrative data from UT and TAMU, the two public flagship institutions, to investigate whether and in what ways the top 10% law modified established high school feeder patterns. Specifically, we first ask whether the share of high schools sending applicants and enrollees has increased and whether the concentration of applicants and enrollees from particular high schools has decreased. Second we ask to what extent applicant and enrollment sending patterns are (1) more expansive geographically; (2) more diverse along socioeconomic lines; and (3) from more diverse high schools under the top 10% admission regime compared with the pre-Hopwood period. Third, using a hazard model, we evaluate whether high school campuses continue sending students in subsequent years after becoming a sending school, and how this persistence has changed after the introduction of the top-10% law. Our results have important policy implications beyond the State of Texas because other states (e.g., Michigan) have begun consideration of a percent plan (Fraser, 2008) and because many states are seeking alternatives to diversify their student bodies along many dimensions, including geographic.

The remainder of this paper proceeds as follows. Section 2 reviews prior studies about high school feeding patterns and college destinations, followed by a brief description of the data in Section 3. Sections 4 and 5 present both methods and answers to the first and second questions about the nature and magnitudes of changes in feeding and sending patterns to UT and TAMU. In Section 5, which addresses the third question, we present the hazard model results. The conclusion discusses the policy implications of our empirical results, and stresses the importance of transparency as a key feature of the top-10% law.

2. Background

A vast literature in sociology and economics shows a strong positive association between students’ socioeconomic background and postsecondary outcomes, including enrollment persistence and completion. Comparatively fewer studies examine variation in high school characteristics and post-secondary outcomes, but several recent studies based on Texas are noteworthy exceptions. For example, Niu, Tienda, & Cortes (2006) demonstrate that graduates from affluent suburban high schools are more likely, and those from schools with large shares of economically disadvantaged students are less likely, to seek admission to selective colleges compared with their statistical counterparts who graduate from typical Texas high schools. Furthermore, these differences in institutional preferences persist among students eligible for automatic admission to any Texas public university of choice. As a partial explanation, Niu, Sullivan, and Tienda (2008) explain that rank-eligible minority students are less likely to know about the provisions of the law than their nonminority counterparts, which results in a potential loss of talented applicants. Furthermore, high schools differ appreciably in their college going traditions, which is an important context for cultivating postsecondary aspirations (Niu and Tienda, 2008).

In addition to individual and family characteristics, claims that the attributes of schools also contribute independently to educational outcomes date back to the controversial 1966 Coleman Report (Coleman, 1990), which documented huge inequities in the resources available in minority dominated schools. Although he argued that economically disadvantaged minority students benefitted educationally from attending integrated schools and classrooms, Coleman was criticized for minimizing the influence of schools on educational outcomes.6 His subsequent work comparing students from public and private high schools showed that Catholic schools, although less generously funded than most private schools, fared better on a range of educational metrics than their counterparts who attended better-endowed schools (Coleman and Hoffer, 1987; Coleman, Hoffer and Kilgore, 1982).

Since Coleman’s pioneering work, relatively few researchers have succeeded in detecting “school effects” on educational outcomes, despite formidable methodological and computational innovations. Two recent studies based on Texas are relevant exceptions. Frost (2007), who uses multi-level modeling to examine whether the socioeconomic and ethno-racial mix of Texas high schools is associated with students’ expectations to graduate from a four-year college, finds that both school socioeconomic level and achievement composition influence students’ college expectations. Niu and Tienda (2008) evaluate variation in college choice sets among a representative sample of Texas high school seniors. They too find that type of high school attended is far more decisive in shaping college choices than students’ academic achievement. Both studies highlight the importance of evaluating whether the top 10% law increased representation at the public flagships from schools populated by low-income and minority students.

The limited body of research about high school sending patterns indicates that well-developed social networks and access to academic resources are decisive in cultivating high post-secondary aspirations, including admission to the most competitive institutions (Frost, 2007). Established alumni loyalties and parents’ institutional affiliations further reinforce feeder patterns across generations by bolstering children’s college choices based on their parents’ legacy status. High school feeder patterns to colleges can embody long-standing regional and community loyalties that have lasting consequences for an institution’s socio-economic and geographic diversity because they can become self-perpetuating. Some scholars (Perna and Titus, 2005) show that certain high schools are better at nurturing feeder legacies that can be leveraged effectively for the benefit of future cohorts of graduates, while other scholars (Wolniak and Engberg, 2007) find that networks between high schools and colleges can strongly influence students’ college aspirations, and hence application and enrollment decisions. Entrenched high school feeder patterns can thus “bound” or limit social influence on the college choice process (McDonough, 1997; Niu and Tienda, 2008).

Established high school feeder patterns also can privilege students from higher socioeconomic backgrounds in ways that simultaneously disadvantage students from lower socioeconomic and underrepresented populations. Martin, Karabel, and Jaquez (2005) examined admission patterns of high schools with long traditions of sending students to the University of California system. As appears to be the case in Texas, they find that a small group of high schools accounted for a disproportionate number of students at the premier UC campuses and that these feeder high schools largely enroll affluent white and Asian students. High schools that serve low-income or predominantly minority students send proportionately and numerically fewer applicants and enrollees to the UC campuses. Their findings are consistent with those of Wolniak and Engberg (2007), who argue that higher education institutions should seek to broaden sending networks from districts whose students are underrepresented on their campuses.

Social class barriers to college access, whether real or self-imposed due to poor understanding of entry requirements and financial aid opportunities, also restrict geographic and economic diversity at selective institutions (Bowen, Kurzweil and Tobin, 2005; Koffman and Tienda, 2008; Niu, et al., 2008). For example, Astin and Oseguera (2004) find that students from the wealthiest families are overrepresented at selective institutions by a two to one margin relative to peers from the poorest families, and that this socioeconomic enrollment gap has grown over time.

Given the entrenched nature of high school feeder patterns, meaningful changes are likely to evolve slowly. Yet, there is suggestive evidence that the top 10% law altered the sending patterns to the public flagships not only by redistributing the applicant pool among public institutions, but also by explicitly allowing rank-eligible students to select their preferred campus (Long and Tienda, 2008b). Because Texas students are qualified for the admission guarantee on a school-specific basis, eligibility is more transparent than in the UC system, where qualification is determined on a statewide basis using a multi-dimensional academic index. In Texas, students need only know their class rank, and school administrators and college counselors need to encourage their top performing students to submit applications—a requirement that often is conducted as part of senior English classes. Thus, the transparency of the admission guarantee for rank-eligible graduates potentially can weaken the social networks that perpetuate the entrenched feeder patterns and consequently broaden geographic, as well as socioeconomic diversity.7 In the remainder of the paper, we investigate whether and how high school sending patterns of applicants and enrollees to the public flagships changed in response to a more transparent admission policy.

3. Data

The University of Texas administrative data for our research comes from two sources. One consists of individual-level applicant data for the years 1990 to 2003 that have been compiled by the Texas Higher Education Opportunity Project (THEOP, www.theop.princeton.edu). We have collapsed these data to the high school level and focus on the number of students attending regular Texas public high schools that apply to and enroll at UT during the observation period.8 The second source is publicly available data from the UT-Austin Office of Admissions Research (OAR) for the years 1996 to 2007. These data contain high-school level information on the number of applicants and enrollees for Texas public high schools that sent one or more enrollees to UT. Because these data lack information on the universe of high schools that sent applications to UT, we use this dataset only to evaluate changes in enrollment. For the overlapping years 1996 to 2003, the THEOP and OAR data contain nearly identical numbers of enrollees per high school. Thus, we only use the OAR data for the years 2004 to 2007. The Texas A&M administrative data is individual-level applicant data for the years 1992 to 2002 compiled by THEOP.

These datasets have been merged with the U.S. Department of Education’s Common Core of Data (CCD), which provides information about several high school attributes of interest on a time-varying basis, including location, enrollment, racial composition, and share of students receiving free- or reduced price lunch. The analysis sample excludes private and alternative high schools, as well as public high schools that lacked a senior class, which is the year college applications are submitted.9 After exclusion restrictions, the analyses universe includes 1,846 unique secondary school campuses for the observation period.

The CCD identifies the “urbanicity” of high schools in eight categories.10 We reduce this categorization into four types:

  • “Urban” = Within the city limits of the principal city of a large urban area (Austin, Corpus Christi, Dallas-Fort Worth-Arlington, El Paso, Houston, or San Antonio).

  • “Suburban” = Within the urban area of these large cities, but not within the city limits.

  • “Rural” = Rural area outside of a Metro or Micropolitan Statistical Area.

  • “Town or Midsize City” = Any area not otherwise captured above, including midsize cities and towns.

We further re-categorize any high school that lies within a 20-mile radius of the center of the large cities as “Suburban.” Thus “Rural” and “Town or Midsize City” only consists of areas outside those radii and outside of these large cities’ Metropolitan Statistical Areas. Given its rapid population growth during the observation period, the “Town or Midsize City” category is of particular interest because it likely includes many new schools.

4. Findings: Have High Schools’ Application and Enrollment Patterns Changed?

As presaged by Montejano’s (2001) early assessment, Figure 1 shows that the share of public Texas high schools sending applicants and enrollees to UT increased over time. The share sending at least one applicant to UT remained stable between 1992 and 1996, a period when affirmative action was permitted, but plummeted in 1997, when the Hopwood decision banned consideration of race in college admissions. The chilling effect of the Hopwood decision on applications proved temporary, however. In 1998, the first year the top 10% law was in force, the share of high schools sending applications rebounded and remained slightly above the pre-Hopwood level through 2001, and then steadily climbed higher through 2003. Enrollment trends followed suit, and the share of high schools sending enrollees continued to climb through 2007. The share of public high schools sending enrollees is 8 to 10 percentage points lower, on average, than the share sending applications, and the annual fluctuations are less pronounced.

Figure 1.

Figure 1

Open in a new tab

Share of Texas Public High Schools Sending at Least One Applicant/Enrollee to UT-Austin or Texas A&M

While the number of sending schools remained higher at TAMU than at UT throughout the period, there is less evidence of increased access at TAMU resulting from the policy change. The share of high schools represented among TAMU applicants rose under affirmative action, sharply fell in 1997 when it was banned, and then modestly increased through 2002 while not returning to its pre-Hopwood peak. The share of high schools represented among TAMU enrollees increased gradually, with little evidence of any spike or change in trend after the uniform admission policy was implemented.

Gini indexes permit a more definitive assessment of whether the applicant and enrollee sending patterns became less concentrated.11 Results shown in Figure 2 reveal clear evidence of a reduced concentration of enrollees at UT under the top 10% policy.12 Between 1990 and 1997, application and enrollment to UT became more concentrated at particular high schools, which squares with earlier claims that a handful of high schools sent disproportionately large numbers of graduates to UT (Montejano, 2001; Tienda and Niu, 2006b; Saenz, 2007). After the top-10% policy was enacted, the trend toward increased concentration of applicants from a relatively few high schools halted, evident in the lower enrollment Gini ratio between 2000 and 2007. The steep fall in 2002 reflects the rescission of a temporary increase in the size of the undergraduate class in 2000 that proved unsustainable because it exceeded the carrying capacity of the university (Tienda and Sullivan, 2009).

Figure 2.

Figure 2

Open in a new tab

Concentration of HS Enrollment and UT and TAMU Applicants and Enrollees Coming from Particular Texas Public High Schools

For TAMU, which drew upon a larger number of sending high schools before the top 10% law, there is no evidence of a reduced concentration of high school feeding patterns. Instead, the College Station campus witnessed an increased concentration of applications after 1997. Furthermore, the high school concentration of enrollment spiked sharply during the first year of the top-10% law (1998) and then fell steadily, converging to the 1997 level. Given the shorter observation period for TAMU, it is not clear whether the decline after 1998 is a reaction to the top-10% law. The law’s apparent disparate impact on high school feeder patterns to UT and TAMU reinforces the prevailing institutional cultures within each campus, namely that TAMU has had more of a historical challenge in diversifying its student body. We consider this disparate impact further in the conclusion.

It is also worth noting that the top-10% law has created a greater admission squeeze on UT than on TAMU (Tienda and Sullivan, 2009). The share of admitted UT students who were in the top-10% of a Texas high school ranged from 41 to 54% between 1998 and 2002, surged to 70% in 2003 and reached 81% in 2008 (Schevitz, 2008; University of Texas, 2008). At TAMU, the share of admitted students who graduated in the top-10% of their Texas high school class rose from 42% to 47% between 1997 and 2002 (THEOP data) and has since receded back to 44% by 2007 (THECB, 2009). UT’s growing enrollment saturation with top-10% students, coupled with less concentrated sending patterns suggests that the characteristics of high schools sending enrollees to UT likely changed. Therefore, we consider whether the sending schools are more diverse in their geographic, ethno-racial composition and socioeconomic characteristics. In the interest of parsimony, we focus on enrollees, but the patterns are quite similar for applicants.13

5. Findings: Have the Characteristics of Sending Schools Changed?

Representation by High School’s Urbanicity

Because of unsubstantiated claims that the top 10% law has increased access to students from sparsely populated regions of the state, we first consider whether feeding patterns have changed by level of urbanization. To address this question, we compute the shares of UT’s enrollees originating from urban, suburban, rural, and other midsize cities or towns as well as the shares of total 12th grade enrollment coming from each area. The ratio of each area’s share of UT enrollees and its share of 12th graders indicates over/under-representation by level of urbanicity. A ratio greater than one indicates over-representation for that category.

Figure 3 shows that suburban and urban high school students have been overrepresented among enrollees in all years, but especially those from suburban areas. The degree of suburban overrepresentation among enrollees increased steadily between 1990 and 1997, and then fell gradually but continuously over the next decade. By contrast, graduates from rural areas and small- and mid-sized cities were severely underrepresented during the first half of the 1990s, and their underrepresentation modestly worsened by 1997. Under the top-10% admission regime, the underrepresentation of these students improved gradually, especially for graduates from rural high schools. By contrast, the Hopwood decision and the top-10% law appear to have lowered access for students from urban high schools, whose relative overrepresentation declined steadily from 1991 to 2006. Thus, the top 10% law appears to have shifted UT’s enrollment representation away from suburban students toward those from rural and small- and mid-sized cities, just as proponents (and critics) have suggested.

Figure 3.

Figure 3

Open in a new tab

Over/Underrepresentation Among Enrollees at UT, by High School’s Urbanicity

Representation by High School Poverty

To address whether the deconcentration of UT’s sending involved greater numbers of students who attended low-income schools with large minority populations, as envisioned by the architects of the top 10% law, we replicate these analyses for several high school strata. We first divide high schools by percentage of their students receiving free- or reduced price lunch (FRPL) which serves as a proxy for student poverty. We then average the high school’s percentage of FRPL students over the period 1990 to 2006, and classify them into four tiers: <20%, 20–40%, 40–60%, and >60%. We repeat this procedure using each high school’s share of students who were black, Hispanic, or Native American.

Figure 4 reports trends in over- and underrepresentation of high schools according to their share of low-income students. Throughout the observation period, students from low-poverty high schools are overrepresented among UT enrollees and students from high schools where more than 20% of their students receive free- or reduced price lunch are underrepresented. Yet after 1998, the overrepresentation of students from low-poverty high schools began a downward trend, as the share of UT enrollment from the highest poverty schools inched upward. Most impressive is the growing representation of students from schools where 40 to 60% of students receive free or reduced lunches.14 These trends provide further evidence that the top 10% law broadened access to students attending less affluent schools.

Figure 4.

Figure 4

Open in a new tab

Over/Underrepresentation Among Enrollees at UT, by High School’s Poverty

Representation by High School Minority Student Share

As demonstrated by many studies, white students enjoyed disproportionate representation at the state’s top public flagship. Figure 5 shows that high schools with less than 40% of their students from members of underrepresented minority groups (URMs) (i.e., black, Hispanic, or Native American) are overrepresented among enrollees at UT, while students from high schools with more than 60% of their student body consisting of URMs were underrepresented throughout the period. Between 1990 and 1997, the high schools with the lowest share of URMs increased their overrepresentation at UT, but this trend reversed after the introduction of the top-10% law. By contrast, the highest-URM schools gained representation among freshman enrollees, particularly after 2002. The rapid convergence in representation between high- and low-URM schools in 2003 reflects the sudden contraction in total undergraduate enrollment following the rescission of the temporary expansion in class size between 2000 and 2002 (see Tienda and Sullivan, 2009).

Figure 5.

Figure 5

Open in a new tab

Over/Underrepresentation Among Enrollees at UT, by High School’s Minority Student Share

Representation by High School Region

To further illustrate the geographic dimensions of less concentrated sending patterns, we examine changes in the regional composition of UT enrollment. It is conceivable that the top-10% law might alter most the application and enrollment behavior of students at high schools from regions of Texas that are distant from the capital because these schools likely had less consistent sending patterns to UT. To examine this proposition, we first classify high schools into the 13 regions defined by the Texas Comptroller, and then collapse the categories into five.15 Figure 6 displays the degree of representation for the five following regions: Alamo (including San Antonio), Capital (including Austin), Gulf Coast (including Houston), Metroplex (including Dallas), and the remaining nine regions combined. For simplicity of exposition, we use a different metric for comparisons – namely the difference (rather than the ratio) between share of UT enrollment and share of 12th grade enrollment for each group of high schools.

Figure 6.

Figure 6

Open in a new tab

Over/Underrepresentation Among Enrollees to UT, by High School’s Region

Representation by High School’s Region

In 1990, the Capital and Gulf Coast regions were greatly over-represented among UT enrollees, which partly reflects the historical feeding patterns facilitated by geographic proximity to the flagship institution. UT enrollment of students from the Alamo and Metroplex regions was comparable to their share of 12th grade enrollment, but the other nine regions were underrepresented at the Austin campus. By contrast, the share of UT enrollment coming from the Capital region in 1990 was three times larger than the region’s share of 12th grade enrollment.

Overrepresentation of schools from the Capital and Gulf Coast regions rose through 1996, but has since fallen steadily. Representation of schools from the Metroplex region hovered around parity over the period, dipping slightly after 2000, but rebounding after 2003. Graduates from Alamo region schools maintained their UT enrollment share at par throughout the period. As the Capital and Gulf Coast region schools gained representation during the early 1990s, the relative shares of students from the other nine regions fell. Collectively, these regions contributed only 19% of UT enrollment in 1996 despite yielding 42 percent of Texas’ high school graduates that year. The tides turned in 1997 when students from these regions increased their representation on the UT campus.

Collectively, these results consistently show increased access to UT from high schools that traditionally have had low representation, namely those serving students from rural, mid-size cities, and regions of the state that previously produced fewer enrollees. Students from schools that serve large numbers of URMs and low-income students also increased their UT enrollment shares after 1997. The widening of access appears to have accelerated in 2003, coincident with the contraction in total enrollment. For the most part, broadened geographic access aligns with the intent of the law and represents a reversal of trends underway before H.B. 588.

Change in Share Enrolling at UT as a Function of High School Characteristics

Because low income and minority-dominated schools are disproportionately located in large urban districts, we evaluate the changes in a regression framework both to assess whether the observed trends are statistically significant and to determine which school characteristics are the most salient when considered jointly. The dependent variable for this analysis is the change in the share of a high school’s graduates enrolling in UT between 1996 and 2007. We restrict the analysis to high schools in existence in both years. Given that the prior figures show pre-Hopwood trends favoring students from suburban, low-poverty, low-minority, and Capital and Gulf Coast region schools, the regression results are conservative estimates of the effect of the top-10% law on high school representation. The results are shown in Table 1.

Table 1.

1996 to 2007 Change in the Share of the High School’s Students Enrolling at UT-Austin

(in percent)
(1) (2) (3) (4)
Urbanicity
 If in suburban area −0.19% [0.37%] 0.66% * [0.34%]
 If in rural area 0.98% *** [0.31%] 1.43% *** [0.33%]
 If in town or mid-sized city 0.70% ** [0.32%] 0.63% * [0.33%]
Percent of students who were black, Hispanic, or Native American −0.93% * [0.49%] 0.94% [0.58%]
Percent of students receiving Free- or Reduced-Price Lunch 5.46% *** [0.95%] 2.70% *** [0.86%]
Region
 If in Capital region (Austin) −3.37% *** [0.80%] −2.93% *** [0.75%]
 If in Gulf Coast region (Houston) −0.81% * [0.48%] −0.47% [0.42%]
 If in Metroplex region (Dallas) −0.20% [0.41%] 0.21% [0.36%]
 If in other region (excl. Alamo, Capital, Gulf Coast, Metroplex) 0.33% [0.37%] −0.03% [0.34%]
Constant −0.87% *** [0.28%] −1.59% *** [0.23%] −0.27% [0.36%] −2.11% *** [0.49%]
Observations 1139 1139 1139 1139
R-Squared 3.6% 9.4% 12.2% 19.6%
Open in a new tab

Standard errors in brackets

***

p<0.01,

**

p<0.05,

*

p<0.1

Column (1) reveals a significant decline in the share of students from urban high schools enrolling at UT (as reflected by the constant). Suburban schools’ enrollment rate declined more than that of urban schools, but the difference is statistically trivial. Conversely, the enrollment rates of students from rural schools and schools in small to mid-size cities significantly increased relative to urban schools. Column (2) shows a significant enrollment gain of students from high-poverty schools, but an enrollment share loss of students attending schools with larger shares of URMs, controlling for the high school’s poverty. Column (3) reveals a significant loss in enrollment share at Capital and Gulf Coast region schools (relative to schools in the Alamo region, as reflected by the constant). UT enrollment rates of high schools in the other nine regions increased relative to the Alamo region, but the difference was not significant.

The final analysis (shown in Column (4)) combines geographic, socioeconomic, and population composition attributes to determine whether the bivariate associations capture covariation among school attributes. Most bivariate relationships shown in columns 1–3 persist, while some do not. Notably, controlling for other characteristics of the school, enrollment rates of students in suburban schools rose relative to urban schools. Once again, there is no evidence that schools in the “other” regions gained representation relative to schools in the Alamo, Gulf Coast, Metroplex, and Capital regions controlling for the schools different economic, demographic, and urbanicity attributes.

6. Findings: Persistence of Sending Schools

Given that the top-10% policy appears to have broadened access to the University of Texas for high schools that traditionally sent fewer students to UT, particularly high-poverty and rural schools, we next consider how long a high school remains a sender of enrollees, conditional on becoming a sender. For this analysis, we have constructed a dataset of “sending spells” where each “spell” represents the duration of the time that the school remains a sender of enrollees to UT. We then evaluate the factors that increase or decrease the “hazard” of the sending spell ending (i.e., the hazard that the high school sends zero enrollees to UT in a given year). The hazard analysis poses one major challenge: for a high school that sent enrollees to UT in 1990 (the first year of our data), we lack information about the number of “sending” years before 1990 (i.e., the data is “left-censored”). As a sensitivity analysis, we conduct the hazard analysis both including and excluding the spells that begin in 1990.

Figure 7, which plots the Kaplan-Meier survival functions, shows that nearly half of all high schools that sent enrollees to UT in 1990 sent enrollees continuously through 2007. The survival sending rate falls more quickly for sending spells that began after 1990, such that within four years less than one-quarter of these spells remain. The higher hazard rate for spells that begin after 1990 is as expected; by definition, these spells originate from high schools that have less attachment to UT as sender schools.

Figure 7.

Figure 7

Open in a new tab

High schools’ peristence in sending enrollees to UT

We next estimate the hazard rate as a function of school characteristics using a Weibull distribution.16 Results reported in Table 2 present coefficients in exponentiated form, where those less than one indicate factors that lower the hazard that the high school’s spell will end, and coefficients larger than one identify factors that raise the hazard rate. Columns (1)-(3) contain all spells, including those “beginning” in 1990. The first set of results show that spells end faster for schools with more poor students and schools in the Metroplex region; however, the hazard rate is lower for larger enrollment schools, those located in suburban areas, schools with more black, Hispanic, or Native American students, and those located in the Alamo, Capital, or Gulf Coast regions. Consistent with Figure 7, the hazard rate is lower for spells that “began” in 1990. Estimates in Column (2) include a dummy variable for 1997 (the year that the Hopwood decision was effective) and another dummy variable that indexes the years when both Hopwood and the top-10% law (H.B. 588) were in force (1998–2007). These results indicate that the hazard of a sending spell ending is significantly lower during the H.B. 588 period, which suggests that the top-10% law may have strengthened persistence in sending rates. The interaction of the H.B. 588 period dummy with other control variables reveals that the higher persistence in sending patterns largely involves urban, high minority, and Gulf Coast high schools.

Table 2.

Survival Analysis: Years that a High School Remains a Sender of Enrollees to UT-Austin
All “Spells” “Spells” beginning after 1990
(1) (2) (3) (4) (5) (6)
Number of enrollees in prior year 0.544*** [0.015] 0.544*** [0.015] 0.543*** [0.015]
12th grade enrollment 0.096*** [0.009] 0.117*** [0.010] 0.110*** [0.010] 0.183*** [0.017] 0.185*** [0.017] 0.167*** [0.016]
Located in suburban area 0.549*** [0.089] 0.531*** [0.083] 0.314*** [0.062] 0.507*** [0.084] 0.475*** [0.080] 0.249*** [0.055]
Located in rural area 0.914 [0.152] 0.935 [0.149] 0.578*** [0.114] 0.877 [0.147] 0.896 [0.153] 0.554*** [0.121]
Located in town or mid-sized city 0.918 [0.147] 0.874 [0.134] 0.599*** [0.115] 0.807 [0.130] 0.779 [0.128] 0.516*** [0.109]
Percent of students who were black, Hispanic, or Native American (URM) 0.797*** [0.061] 0.771*** [0.056] 0.969 [0.091] 0.811*** [0.061] 0.789*** [0.060] 0.938 [0.094]
Percent of students receiving Free- or Reduced-Price Lunch (FRPL) 1.495*** [0.114] 1.519*** [0.111] 1.344*** [0.125] 1.405*** [0.105] 1.447*** [0.111] 1.329*** [0.132]
Located in Alamo region (San Antonio) 0.666*** [0.103] 0.679*** [0.100] 0.823 [0.150] 0.743* [0.114] 0.742* [0.116] 0.883 [0.172]
Located in Capital region (Austin) 0.213*** [0.041] 0.234*** [0.044] 0.273*** [0.065] 0.333*** [0.066] 0.322*** [0.066] 0.364*** [0.096]
Located in Gulf Coast region (Houston) 0.702** [0.101] 0.726** [0.100] 0.942 [0.163] 0.703** [0.101] 0.719** [0.105] 1.024 [0.190]
Located in Metroplex region (Dallas) 1.281** [0.137] 1.285** [0.132] 1.211 [0.166] 1.223* [0.129] 1.253** [0.135] 1.177 [0.174]
Hopwood (1997) 0.824* [0.085] 0.801** [0.082] 0.969 [0.101] 0.958 [0.100]
H.B. 588 Years (1998+) 0.450*** [0.025] 0.215*** [0.048] 0.582*** [0.032] 0.291*** [0.069]
HB588 * Suburban 2.671*** [0.652] 3.127*** [0.807]
HB588 * Rural 2.389*** [0.563] 2.243*** [0.555]
HB588 * Other Area 1.971*** [0.463] 2.024*** [0.499]
HB588 * Percent URM 0.674*** [0.073] 0.745*** [0.084]
HB588 * Percent FRPL 1.238** [0.133] 1.150 [0.129]
HB588 * Alamo 0.662* [0.140] 0.677* [0.149]
HB588 * Capital 0.676 [0.196] 0.734 [0.224]
HB588 * Gulf Coast 0.631** [0.131] 0.540*** [0.116]
HB588 * Metroplex 1.069 [0.164] 1.059 [0.171]
Spell Began in 1990 0.674*** [0.059] 0.387*** [0.035] 0.348*** [0.033]
Constant 0.119*** [0.019] 0.202*** [0.032] 0.292*** [0.055] 0.195*** [0.031] 0.257*** [0.042] 0.363*** [0.073]
Ln (p) 1.706*** [0.037] 1.777*** [0.037] 1.840*** [0.038] 2.084*** [0.040] 2.161*** [0.040] 2.225*** [0.041]
Number of spells 2,690 2,690 2,690 2,588 2,588 2,588
Open in a new tab

Standard errors in brackets

***

p<0.01,

**

p<0.05,

*

p<0.1

To address left censoring, Columns (4)-(6) repeat these analysis using only spells that began after 1990. The results are virtually unchanged. One advantage of restricting the analysis in this way is that we can incorporate into the analysis the number of students who enrolled at UT from the prior year’s graduates for each high school. Even controlling for the high school’s 12th grade enrollment, the more students from the prior class are represented at UT, the lower the hazard that the high school will send no enrollees in the current year. This result confirms that high schools with a strong feeding history self-perpetuate UT representation by influencing the enrollment behavior of subsequent cohorts of students. Combined with the finding that the top-10% law has reduced the hazard rate (and thereby increased sending persistence), our results suggest that the law’s impact on sending patterns could magnify over time as the high school accumulates a sending history. Thus, to some extent, the decreasing concentration of enrollment at UT from particular high schools, which gained momentum beginning in 2003 (five years after the beginning of the top-10% law), could partly result from this dynamic cross-cohort relationship.

7. Conclusion

In this paper, we sought to evaluate whether a state-level college admission policy that targets high schools and guarantees admission to a designated percentage of graduates can both weaken entrenched sending patterns and increase geographic and socioeconomic diversity. We find that the number and share of high schools represented among University of Texas applicants and enrollees rose in certain areas after the introduction of the top-10% law, reversing a pre-policy trend towards more concentration. It is unclear whether the enrollment and sending patterns have been permanently changed, however. We find no evidence for similar effects at Texas A&M University, the State’s first public university, but there are many reasons for this result. First, TAMU drew from a larger number of high schools than UT even before H.B.588 went into effect. Second, TAMU’s sending school patterns may be more entrenched, even if less concentrated, owing to the institution’s long-standing presence in nonmetropolitan communities (through its Agriculture Extension Service Offices across the state).17 These longstanding institutional ties may have helped to forge feeder school legacies long ago, whereas UT-Austin is just now making inroads in these communities. Third, although situated within a triangle of Interstate highways that link three major cities—Houston, Dallas and Austin—TAMU’s nonmetropolitan location (College Station) may be less appealing to students than metropolitan Austin.

Another major finding is that the new policy increased the representation at UT-Austin of students from high-poverty schools, those with greater shares of minority students, those located in rural areas as well as and small- and mid-sized cities, and schools located in regions that traditionally were underrepresented at UT. Further, we show that the number of consecutive years that high schools sent enrollees to UT increased after the introduction of the top-10% law. Thus, schools that become new senders of enrollees to UT are likely to persist in sending enrollees for longer periods of time, potentially creating a historical legacy that will increase the likelihood of enrollment by future cohorts of students at these new sending schools. Our results suggest there is potential value in strengthening institutional linkages between new sending schools and public universities as a strategy to deepen college-going traditions in the new sending schools—actions that may become even more important in the future if the top 10% law is rescinded.

To some extent, our finding that the number of high schools represented among UT’s applicant and enrollee pools increased under the top 10% regime compared to the pre-Hopwood era is somewhat surprising because even before H.B. 588 became law, virtually all applicants who graduated in the top decile of their high school class were admitted to UT-Austin (Long and Tienda, 2008a). Presumably, many seniors who ranked highly in their class failed to apply because of the opaqueness of UT’s admissions policy; as is the case at most institutions, students have no way of knowing whether they qualify for admission or the likelihood of being admitted. This opaqueness would be acute for students at high schools with low sending rates to UT – a student at such a high school would not have the experience of seeing their older peers’ application results. Thus, the apparent increases in access may be due, in part, to the rendering of an opaque de facto policy that admitted nearly all top 10% students to a transparent de jure policy that clearly stipulated the criteria for automatic admission. Not only did this admission policy change influence the number of admitted and enrolled students to UT, but it also diversified their geographic and socioeconomic origins, which is consistent with Irma Rangel’s vision when crafting the law.

The consistency of the law’s provisions guaranteeing access to all public universities and the transparency of the uniform merit criterion applied to all Texas public school students differentiates it from the percent plans in California and Florida that do not guarantee admission to all public universities in the state as well all the myriad variants of “full file review” currently used by public and private institutions which are not transparent.

Our results have broad implications for public policies where simplicity and transparency may matter for broadening access to government services or for spotlighting pervasive inequities in other public domains. For example, Currie’s (2004) research shows that take-up rates of various programs, such as 401k plans and Medicare, rise when participants are automatically enrolled by their employer or through their participation in other social programs such as welfare. Additionally, program complexity and lack of information on the part of potential participants has been shown to lower participation in public programs such as welfare, Medicare, or the State Children’s Health Insurance Program (Kleven and Kopczuk, 2008; Aizer, 2007; Bansak and Raphael, 2007), further suggesting that policy transparency (and simplicity) may raise participation rates. Finally, the transparency provided by the No Child Left Behind Act, which requires schools and districts to report separate assessment results for demographic subgroups, has allowed researchers and advocates to document both progress and persistence of achievement gaps between these groups.

Acknowledgments

This research was supported by grants from the Ford, Mellon and Hewlett Foundations and NSF (GRANT # SES-0350990). We gratefully acknowledge institutional support from Princeton University’s Office of Population Research (NICHD Grant # R24 H0047879). We are grateful to Danielle Fumia for outstanding research assistance and to Dawn Koffman and Milagros Nores for helpful suggestions.

Footnotes

1

For anecdotal evidence, see the 60 Minutes episode, “Is The ‘Top 10’ Plan Unfair?”, which aired in October 2004.

2

These estimates are based on tabulations from the THEOP administrative files. Similar patterns obtain for admittee and enrollee pools, except that the absolute numbers are smaller. Over the same period, the number of Texas high schools represented among admittees increased from 635 to 773 at UT, and from 787 to 900 at TAMU. Among enrollees at UT, 547 and 679 high schools were represented in 1992 and 2002, respectively. At TAMU, the enrollee pools in 1992 and 2002 represented 690 and 823 schools, respectively. Historically, TAMU has drawn from more rural populations compared with UT partly because of its land grant mission, and partly because of its location outside of a major city.

3

A notable exception is the temporary increase in the size of the freshman classes in effect from 2000 to 2002, which was an attempt to offset the growing numbers of applicants eligible for automatic admission.

4

Forest (2002) examines the impact of using geographic criteria to diversify TAMU, but his data are only for 1998, the first year the law took effect, and rely only on TEA data, which lack information about actual grades, class rank, or student test scores.

5

For the second consecutive legislative session, in 2007 the Texas legislature considered and scuttled a bill to cap the number of students granted automatic admission at 50 percent, which is comparable to the share of top 10% graduates in 1996, in order to maintain institutional flexibility in shaping their high freshmen classes. The top 10% law will likely be re-debated in 2009, when the Texas legislature reconvenes. The State legislature meets in odd numbered years, except when special sessions are called.

6

Despite polarized scholarly debate about Coleman’s research about school inequality, a more balanced interpretation of his findings is that, compared with family socioeconomic background, school attributes exert smaller influences on educational outcomes.

7

As several papers have demonstrated, an admission guarantee does not ensure enrollment for low–income students, hence knowledge of and receipt of adequate financial aid packages is necessary as well (see Niu, et al., 2008).

8

We exclude alternative high schools and private high schools for different reasons. Alternative high schools have very low college sending programs, as many focus on behavioral problems and vocational programs. The CCD data lack information about private schools, therefore we cannot append comparable attributes for our comparisons. Only one private high school is included in the top 28 feeder schools identified by Tienda and Niu (2006b).

9

Not all public high schools include senior classes; some are separated according to class standing, with one notable large school that only includes seniors.

10

The CCD changed their location coding system for the 2007-07 school year. We base the location code for high schools in this year based on the most recent data from prior years.

11

Gini indexes range from zero to one. A value of zero indicates perfect equality of high schools in terms of their enrollment (or admittees or enrollees). A value of one would indicate that all of the state’s enrollment (or admittees or enrollees) are located in a single high school.

12

The UT-Applicant time-series only extends through 2003; because the OAR data only includes high schools that sent at least one enrollee to UT, therefore it cannot be used to compute Gini indexes for applicants after 2003. For all subsequent analysis beginning with Figure 2, high schools are weighted by their 12th grade enrollment.

13

These results are available from the lead author upon request. Because TAMU did not witness appreciable deconcentration in sending patterns or saturation with applicants whose admission was guaranteed, we do not included TAMU in these analyses.

14

We find a greater increase in representation of high poverty schools among enrollees than among applicants (through 2003). One plausible explanation is the efficacy of the Longhorn Opportunity Scholarships in permitting rank-eligible students from schools with historically low representation at UT to enroll. The Longhorn Opportunity program is targeted to urban public schools with high concentrations of low income students and low college-going traditions (Domina, 2007). The program involves targeted outreach to the designated schools, which simultaneously raises awareness of the admission guarantee and the prospects of a tuition subsidy. An additional interpretation is that the top-10% plan continued to crowd out students from affluent high schools who did not qualify for automatic admission (Niu and Tienda, 2008).

16

Because we may have multiple spells from single high schools, we incorporated high school-level shared frailty into the model. The incorporation of this shared frailty is analogous to incorporating a random effect in an analysis of panel data.

17

The institution’s full name is Texas Agricultural and Mechanical University, which has since been shortened to “Texas A&M,” or TAMU. To what extent the “agricultural and mechanical” legacy its contemporary appeal has not been directly investigated, however.

Contributor Information

Mark C. Long, Email: marklong@u.washington.edu, University of Washington, 206-543-3787.

Victor B. Saenz, Email: vsaenz@mail.utexas.edu, University of Texas at Austin, 512-475-8585.

Marta Tienda, Email: tienda@princeton.edu, Princeton University, 609-258-5808.

References

  1. Aizer Anna. Public Health Insurance, Program Take-Up, and Child Health. Review of Economics and Statistics. 2007;89:400–15. [Google Scholar]
  2. Astin Alexander W, Oseguera Leticia. The Declining “Equity” of American Higher Education. Review of Higher Education. 2004;27:321–41. [Google Scholar]
  3. Bansak Cynthia, Raphael Steven. The Effects of State Policy Design Features on Take-Up and Crowd-Out Rates for the State Children’s Health Insurance Program. Journal of Policy Analysis and Management. 2007;26:149–75. doi: 10.1002/pam.20231. [DOI] [PubMed] [Google Scholar]
  4. Bowen William G, Kurzweil Martin, Tobin Eugene. Equity and Excellence in American Higher Education. Charlottesville, VA: University of Virginia Press; 2005. [Google Scholar]
  5. Coleman JamesS. A Brief Summary of the Coleman Report. In: Coleman James S, Boulder CO., editors. Equality and Achievement in Education. San Francisco, CA; and London: Westview Press; 1990. [Google Scholar]
  6. Coleman James S, Hoffer Thomas. Public and Private High Schools: The Impact of Communities. New York: Basic Books; 1987. [Google Scholar]
  7. Coleman James S, Hoffer Thomas, Kilgore Sally. High School Achievement: Public, Catholic, and Private Schools Compared. New York: Basic Books; 1982. [Google Scholar]
  8. Currie Janet. NBER Working Paper: 10488. National Bureau of Economic Research; 2004. The Take Up of Social Benefits. ( http://www.nber.org/papers/w10488.pdf) [Google Scholar]
  9. Domina Thurston. Higher Education Policy as Secondary School Reform: Texas Public High Schools After “Hopwood”. Educational Evaluation and Policy Analysis. 2007;29:200–17. [Google Scholar]
  10. Forest Benjamin. Hidden Segregation? The Limits of Geographically Based Affirmative Action. Political Geography. 2002;21:855–880. [Google Scholar]
  11. Fraser Kelly. State Lawmaker Wants Ten-Percent Rule: Automatic College Admission Policy has been Implemented in Texas, California, and Florida. Michigan Daily. 2008 Apr 14; [Google Scholar]
  12. Frost Michelle B. Texas Students’ College Expectations: Does High School Racial Composition Matter? Sociology of Education. 2007;80:43–66. [Google Scholar]
  13. Giavanola Anouck. Women’s Legal History Project. Stanford University; Palo Alto, CA: 2005. Irma Rangel. Retrieved January 28, 2008 ( http://womenslegalhistory.stanford.edu/papers05/Rangel_bio_Giovandola. [Google Scholar]
  14. Horn Catherine L, Flores Stella M. The Civil Rights Project. Harvard University; 2003. Percent Plans in College Admissions: A Comparative Analysis of Three States’ Experiences. ( http://www.civilrightsproject.harvard.edu/research/affirmativeaction/tristate.pdf) [Google Scholar]
  15. Hughes Polly R, Tresaugue Matthew. Small-Town GOP Behind Survival of Top 10% Rule; Republicans from Rural Cities Say Law Essential for their Students. The Houston Chronicle. 2007 May 30; B:1 Metfront. [Google Scholar]
  16. Jaschik Scott. The 10% Fight is Back. Inside Higher Education. 2009 Jan 12;
  17. Kain John F, O’Brien Daniel M. “Hopwood and the Top 10% Law: How they have Effected College Enrollment Decisions of Texas High School Graduates.” Working Paper 26. Cecil and Ida Green Center for the Study of Science and Society, The University of Texas; Dallas: 2003. ( http://www.utdallas.edu/research/tsp/pdfpapers/paper26.pdf) [Google Scholar]
  18. Kleven Henrik J, Kopczuk Wojciech. Transfer Program Complexity and the Take Up of Social Benefits. Working Draft. 2008 ( http://www2.warwick.ac.uk/fac/soc/economics/news/forums/conferences/peuk/complexity.2.pdf )
  19. Koffman Dawn, Tienda Marta. Missing in Application: The Texas Top 10% Law and Campus Socioeconomic Diversity. Working Paper Texas Higher Education Opportunity Project 2008 . ( http://theop.princeton.edu/reports/wp/ApplicantSocialClass.pdf)
  20. Long Mark C, Tienda Marta. Winners and Losers: Changes in Texas University Admissions Post-Hopwood. Educational Evaluation and Policy Analysis. 2008a;30:255–80. doi: 10.3102/0162373708321384. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Long Mark C, Tienda Marta. Changes in Texas Universities’ Applicant Pools After the Hopwood Decision. Paper presented at Annual Meetings of the American Education Research Association; April 2008; New York, NY. 2008b. ( http://theop.princeton.edu/reports/wp/Applicant_Pools_Post_Hopwood.pdf) [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Martin Isaac, Karabel Jerome, Jaquez Sean W. High School Segregation and Access to the University of California. Educational Policy. 2005;19:308–30. [Google Scholar]
  23. McDonough Patricia M. Choosing Colleges: How Social Class and Schools Structure Opportunity. Albany, NY: State University of New York Press; 1997. [Google Scholar]
  24. Monastersky Richard. Texas House Rejects Changes in Top-10-Percent Plan for Admissions. Chronicle of Higher Education. 2007;53:A21. [Google Scholar]
  25. Montejano David. Access to the University of Texas at Austin and the Ten Percent Plan: A Three-Year Assessment (Admissions Brief) Report. University of Texas; Austin: 2001. ( http://www.utexas.edu/student/research/reports/admissions/Montejanopaper.htm) [Google Scholar]
  26. Niu Sunny X, Sullivan Teresa A, Tienda Marta. Minority Talent Loss and the Texas Top 10 Percent Law. Social Science Quarterly. 2008;89:831–45. doi: 10.1111/j.1540-6237.2008.00586.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Niu Sunny X, Tienda Marta. Minority Student Academic Performance Under the Uniform Admission Law: Evidence from the University of Texas at Austin. Working Paper Texas Higher Education Opportunity Project. 2008 doi: 10.3102/0162373709360063. ( http://theop.princeton.edu/reports/wp/MinorityAcademicPerformance062008.pdf. [DOI] [PMC free article] [PubMed]
  28. Niu Sunny X, Tienda Marta, Cortes Kalena. College Selectivity and the Texas Top 10% Law. Economics of Education Review. 2006;25:259–72. [Google Scholar]
  29. Paredes Raymund A. Testimony on the Top Ten Percent Law: Raymund A. Paredes, Commissioner on Higher Education. Hearing of the Texas State Senate Subcommittee on Higher Education. 2006 http://www.thecb.state.tx.us/commissioner/testimonytop10.pdf.
  30. Perna Laura, Titus Marvin. The Relationship between Parental Involvement as Social Capital and College Enrollment: An Examination of Racial/Ethnic Group Differences. Journal of Higher Education. 2005;76:485–518. [Google Scholar]
  31. Root Jay. UT Pushes Lawmakers to Modify ‘Top 10% Rule’. Houston Chronicle. 2009 Jan 7; [Google Scholar]
  32. Saenz Victor. Democratization of Access: How the Top Ten Percent Law has Affected the Geographic and Socio-Economic Diversity of Texas High School Feeders to UT-Austin. Paper presented at the Civil Rights Project Research Conference; October 2007; Austin, TX. 2007. [Google Scholar]
  33. Schevitz Tanya. The San Francisco Chronicle (California) Main News; 2008. Texas Chancellor Selected to Lead UC; p. A1. [Google Scholar]
  34. Texas Higher Education Coordinating Board [THECB] First-Time Undergraduate Applicant, Acceptance, and Enrollment Information for Summer/Fall 2007. 2009 ( http://www.txhighereddata.org/interactive/AppAccEnr.cfm)
  35. Tienda Marta, Niu Sunny X. Capitalizing on Segregation, Pretending Neutrality: College Admissions and the Texas Top 10% Law. American Law and Economics Review. 2006a;8:312–46. [Google Scholar]
  36. Tienda Marta, Niu Sunny X. Flagships, Feeders, and the Texas Top 10% Law: A Test of the “Brain Drain” Hypothesis. Journal of Higher Education. 2006b;77:712–39. [Google Scholar]
  37. Tienda Marta, Sullivan Teresa A. The Promise and Peril of the Texas Uniform Admission Law. In: Hall Martin, Krislov Marvin, Featherman David L., editors. the Next Twenty Five Years? Affirmative Action and Higher Education in the United States and South Africa. Ann Arbor: University of Michigan Press; 2009. forthcoming. [Google Scholar]
  38. University of Texas, Office of Admission Research [UT-OAR] “An Investigation into Rural High School Representation in Entering Freshman Classes at the University of Texas at Austin, Summer/Fall Classes of 1996–2007.”. 2008 ( http://www.utexas.edu/student/admissions/research/RuralSchoolStudy-96-07.pdf)
  39. University of Texas, Office of President. “The Top 10% Law and Its Impact on the University of Texas at Austin.” Report on the Top 10% Law. Austin, TX: University of Texas; 2008. [Google Scholar]
  40. Western Interstate Commission for Higher Education [WICHE] Knocking at the College Door: Projections of High School Graduates by State, Income, and race/ethnicity, 1992 to 2022. Boulder, CO: Western Interstate Commission for Higher Education; 2008. [Google Scholar]
  41. Wolniak Gregory C, Engberg Mark E. The Effects of High School Feeder Networks on College Enrollment. Review of Higher Education. 2007;31:27–53. [Google Scholar]

RESOURCES