- Shopping Bag ( 0 items )
In College Choices, Caroline Hoxby and a distinguished group of economists show how students and their families really make college decisions—how they respond to financial aid options, how peer relationships figure in the decision-making process, and even whether they need mentoring to get through the admissions process. Students of all sorts are considered—from poor students, who may struggle with applications and whether to continue on to college, to high aptitude students who are offered "free rides" at elite schools. College Choices utilizes the best methods and latest data to analyze the college decision-making process, while explaining how changes in aid and admissions practices inform those decisions as well.
Sarah E. Turner
More students are attending college than ever before and the labor market rewards to completing a college degree have increased appreciably over the last quarter century. Yet, the rise in the incentives for collegiate completion has not been accompanied by an increase in the share of students making the transition from college enrollment to college completion. Among individuals aged twenty-three in 1970, 23 percent of high school graduates had completed a BA degree, while about 51 percent had enrolled in college for some period since high school graduation. For the same age group in 1999, the share of high school graduates who had enrolled in college at some point rose substantially, to 67 percent, while the share receiving a BA degree rose only slightly, to 24 percent of the cohort. Thus, for college participants measured in their early twenties, completion rates fell by more than 25 percent over this interval. Completion rates measured at older ages are closer to stagnant, implying an overall increase in the time to degree.
It is the combination of collegiate attainment and time to degree that determines the overall supply of workers with college-level skills. The time it takes to complete a degree is an important economic variable in its own right. Delay in degree attainment implicitly lowers the supply of skilled workers to the economy. Moreover, even if individuals receive some consumption benefit by extending their time in college beyond the four-year norm, the public cost is sizable given the high degree of subsidy from state and federal sources. Implicitly, the opportunity cost of extended time to degree (in the absence of perfect elasticity of supply in the collegiate market) is that other students may be denied college opportunities.
That a college education is more important now than ever is certainly cliché, though it is borne out by the overall increase in the college wage premium. The value of a college degree in the labor force has increased substantially, rising from a premium over a high school degree of about 40 percent in 1980 to over 65 percent two decades later. Reduced growth in the supply of college-educated workers may hamper long-term increases in productivity while also increasing the degree of inequality in earnings. How the higher education market transforms student enrollment into collegiate attainment, including degrees conferred, is fundamental to understanding the determinants of the supply of college-educated workers.
It is surprising that collegiate attainment and time to degree have not received more attention. With few exceptions, recent discussions in policy circles have focused on questions related to access, loosely defined as the extent to which individuals from different circumstances enroll in college, to the near exclusion of questions of attainment. Emphasis on vaguely de- fined notions of "collegiate access and affordability" in public discourse has diverted attention from the monitoring of outcomes, such as courses completed and degrees awarded. Enrollment rates are, of course, an important measure of college entry, but they do not provide a measure of the degree to which students and colleges are able to transfer time and resources to completed courses, years of attainment, or degrees earned. These outcomes are measures of human capital acquired and, while necessarily somewhat inexact, they are indicators of the addition to the stock of skills available to the labor force. Degree and credit outcomes register that a student completed a certain path of study with proficiency, while enrollment measures indicate only transitory participation. That the economic return to a BA degree has risen more rapidly than the premium afforded to "some college" is but one indication of the importance of degree attainment.
It is important to ask why many education analysts (including economists) focus on the enrollment measure, which is an indicator of potential investment, rather than on degrees or credits, which measure additions to human capital stock. One explanation is that enrollment is simply much easier to track than outcomes, such as credits earned.
Yet enrollment per se does not capture how individuals, along with colleges and universities, convert "participation" to outcomes such as BA degrees or course credits. That there may be substantial increases over time in the relative enrollment among individuals from poor families or racial minorities need not imply a narrowing in the difference between these groups in collegiate attainment. It is these differences in attainment, not in enrollment, that ultimately affect the distribution of earnings.
The objective of this analysis is to document the changing relationship between college enrollment and college completion, to assess the factors responsible for these shifts, and to consider their implications. In doing so, this analysis sets a new direction for higher education research by documenting the gap between enrollment rates and completions and identifying the universe of possible explanations. The first section considers the measurement of college enrollment and college completion, focusing on the intersection of results from a range of different data sources. The second section sets out a basic framework for analysis, starting with the human capital investment model, and outlines explanations for why individuals who begin college do not complete it or complete it in an extended period of time. In the third section, I provide empirical evidence distinguishing the explanatory role of these various factors. The concluding section summarizes the challenges for future research, as well as suggesting some implications for policy and data collection.
If there is one overriding policy conclusion, it is that the traditional focus of economists and policy analysts on the paired concepts of "enrollment" and "access" is insufficient to insure the supply of college-educated workers needed to meet demand, to reduce income inequality, and to narrow intergenerational differences in education and earnings.
Explaining why completion rates have decreased for those in their early twenties and why time to degree has increased rests on understanding the decisions of individuals to invest in college beyond their initial enrollment. Of particular concern is whether characteristics of today's marginal students, those who might not have started college in previous periods, are systematically different in terms of income or achievement from students beginning college in previous years. Changes over time in the academic preparedness of the marginal student may also reduce completion and increase time to degree. Financial constraints, combined with imperfect access to capital markets, are one demand-side force potentially reducing completion and extending time to degree. Because policy implications associated with credit constraints are dramatically different than those associated with selection effects, considerable care is warranted in distinguishing empirically between these two. Beyond demand-side factors, expansion on the supply-side of the market has been dominated by growth of community colleges and institutions with relatively low resources per student; as such, these institutions are able to contribute less to college completion than are institutions with greater resources per student or more upper-level courses. Public policies, including federal programs such as Pell grants and direct state appropriations to higher education, are not well-targeted and often do not increase opportunities for academically well-prepared students to complete four-year programs.
1.1 The Relationship Between College Enrollment and Collegiate Attainment
The measurement of college enrollment, college participation, and college completion is fundamental to this analysis, but the definition of these variables is often given too little attention. First, college enrollment is inherently a flow variable, representing the number of students participating at a given educational level at a single point in time. College enrollment can be measured from data tabulated by colleges and universities (in which case the age of the enrolled students is often unknown) or it can be tabulated through survey data, including the census, the Current Population Survey (CPS), or other sources, capturing what an individual is doing at a specific point in time. Collegiate attainment is, on the other hand, a stock variable- measuring the sum of education acquired by a given point in time. The metric for measuring collegiate attainment includes measures of credits, years completed, or degrees awarded; implicitly, the defining feature of these variables is that they are nonrevocable. The most general stock measure is "college participation," indicating that an individual completed at least some college.
In this paper, college completion is used to denote the receipt of a four-year baccalaureate degree, though one might identify other types of completion in the undergraduate pipeline, such as receiving the associate degree. Linking initial college enrollment and degree receipt is time to degree. Following the rather considerable literature analyzing time to degree at the PhD level, total time to degree is the gross difference between data at BA completion and initial enrollment, while the net measured or elapsed time to degree captures the calendar period in which a student is enrolled. For any birth cohort, time to degree is an inherently truncated variable as students continue to receive degrees at late ages. Calculation of time to degree from microdata may follow two approaches. First, longitudinal data, such as the National Longitudinal Survey of Youth (NLSY), record the year of degree receipt. Alternatively, repeated cross sections, such as the CPS, afford the opportunity to examine how the educational attainment of a birth cohort changes over time.
In each year, recent high school graduates form the "basic" pool of potential college students, and the fraction of these students who enter college define the "traditional" college enrollment rate. Shown in figure 1.1, the enrollment rate of this group surged in the late 1960s (for men, partly in response to the Vietnam war), and it then stagnated in the 1970s. Between the late 1960s and the mid-1970s, enrollment rates for men and women converged, with the relative decline in enrollment more muted for women than for men over this interval. Since 1980, the rise in the enrollment rate of recent high school graduates has been consistent, and the enrollment rate is now near 65 percent, relative to about 50 percent in 1980.
Collegiate attainment is a function of both initial enrollment rates and the transition of the cohort through the education pipeline. Collegiate attainment, measured for a cohort, is also inherently a truncated variable. A birth cohort measured at age thirty will have had more of an opportunity to acquire education than a birth cohort measured at age twenty-three. Yet the timing of educational attainment is also an economic variable, as individuals acquiring education at relatively young ages will have more years to accrue the returns to the skills they have acquired. By near tautology, increased college enrollment rates of recent high school graduates translate to increases in the fraction of a cohort attaining some college.
Figure 1.2 presents a snapshot of the educational attainment of young adults and shows the proportion completing college and the proportion with any collegiate participation at the age of twenty-three from 1968 to 2000. (The data are presented for birth cohorts from 1945 to 1977, which is analogous to the 1968 to 2000 years of observation.) While participation rises in much the same pattern visible in figure 1.1, the change in the proportion with a college degree is far more muted. There is little visible rise in the share completing college in the birth cohorts born after 1960, in spite of the quite visible increase in participation. Overall, the average annual increase in the college participation rate is 1.1 percent, while the increase in college completion is a more modest 0.7 percent. Beyond the aggregate picture, the data suggest three distinct regimes, with the latest period marking the most substantial divergence between enrollment rates and completion rates. First, for the early cohorts born between 1945 and 1952 (equivalently the children of the baby boom and the college students of the Vietnam era), college enrollment rates and college completion rates both increased sharply for cohorts measured at age twenty-three, with college completion increasing by about 35 percent and college enrollment by about 37 percent over this interval. A reversal followed, with absolute declines in enrollment and completion between the 1952 and 1958 cohorts (those cohorts aged twenty-three between 1975 and 1981), and the relative decline in college completion (about 13 percent) was somewhat larger than the relative decline in enrollment rates (about 18 percent). Then, from the 1958 cohort on, college enrollment increased markedly, surpassing the 1952 local maximum by 10 percentage points by the time those born in the late 1970s reached the age of twenty-three.
Thinking about the difference between enrollment rates and completion rates as a difference in levels conveys much of the same information and also illustrates the widening gap between enrollment rates and completion in recent birth cohorts. Among those born in 1957 and aged twenty-three in 1980, the expected difference between enrollment and BA completion among high school graduates was about 27 percentage points; by 2000, the gap was 36 percentage points for the cohort aged twenty-three (born in 1977). It follows that the college completion rate (the share of those with some college receiving a degree) decreased from nearly 40 percent to about 34 percent, with this trend shown in the bottom panel of figure 1.2.
Turning to the same trends in college participation and completion for demographic subgroups, figure 1.3 shows the trends for men and women and figure 1.4 shows the trends for blacks and whites. Gains in college participation are marked for blacks, rising at an average annual rate of 2.5 percent, though these gains are not replicated in the completion measure. Men and women display about the same modest overall decline in completion rates, but for men this is against a backdrop of stagnant college participation, while college participation has been rising for women. For each subgroup, completion rates decline over the entire interval, though the decent is strikingly larger for blacks than for those in other ethnic groups.
The observation of individuals at age twenty-three is a truncated picture of completion; changes in time to degree and the age structure of enrollment also need to be considered. To provide a firmer understanding of how these measures of collegiate attainment change over time, figure 1.5 shows college completion and college enrollment over time for different age levels. Most striking is the divergence between the top panel, showing participation, and the bottom panel, showing completion. For the most part, students who will participate in the collegiate system have had at least some college by age twenty-two, as the share recording some college for each birth cohort at this age is nearly identical to the share with some college for age thirty. It is in the bottom panel showing college completion where we see substantial divergence by time and by age. For all cohorts there are gains in BA completion by age, but these differences become particularly pronounced after the 1955 birth cohort, where the share of twenty-two-year-olds with a BA degree actually declines while degree receipt increases at older ages, particularly over twenty-five. That few of the students beyond age twenty-two are new participants provides an indication that either the duration of enrollment required to receive a BA has increased or more students complete their degrees after a series of spells of discontinuous study. Thus, for students receiving BA degrees between ages twenty-eight and thirty, the total time to degree likely exceeds ten years.
Excerpted from College Choices Copyright © 2004 by National Bureau of Economic Research. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.
1. Going to College and Finishing College: Explaining Different Educational Outcomes - Sarah E. Turner
Comment: Christopher Taber
2. The New Merit Aid - Susan Dynarski
Comment: Charles Clotfelter
3. The Impact of Federal Tax Credits for Higher Education Expenses - Bridget Terry Long
Comment: Michael McPherson
4. Education Savings Incentives and Household Saving: Evidence from the 2000 TIAA-CREF Survey of Participant Finances - Jennifer Ma
Comment: Harvey S. Rosen
5. How Financial Aid Affects Persistence - Eric Bettinger
Comment: Jonathan Guryan
6. Do and Should Financial Aid Packages Affect Students' College Choices? - Christopher Avery and Caroline M. Hoxby
Comment: Michael Rothschild
7. Resident and Nonresident Tuition and Enrollment at Flagship State Universities - Michael J. Rizzo and Ronald G. Ehrenberg
Comment: Michelle J. White
8. Student Perceptions of College Opportunities: The Boston COACH Program - Christopher Avery and Thomas J. Kane
Comment: Bruce Sacerdote
9. Peer Effects in Higher Education - Gordon C. Winston and David J. Zimmerman
Comment: Thomas S. Dee