The Earnings Effect of Education at Community Colleges

Beginning of article

I. INTRODUCTION

Since the early 1980s, the earnings of workers with a postsecondary education have grown substantially, relative to those of their high school educated peers (Autor, Katz, and Kearney, 2005). Economists explain the bulk of this increase in relative earnings as the result of shifts in the structure of demand in the labor market that have favored workers with more skill (e.g., Levy and Murnane, 1992; Bound and Johnson, 1992; Autor, Katz, and Kearney, 2005). This explanation resonates with industry groups and policy makers, whose response has been to demand accountability in secondary education and to encourage high school students (and workers looking to retool) to attend college.

While they receive relatively little attention, community colleges are one potentially important set of institutions providing access to post-secondary education, among both recent high school graduates, and older workers attempting to upgrade their skills. (1) Nearly 40% of all students enrolled in postsecondary education are at 2-yr institutions. (2) Furthermore, enrollment at 2-yr institutions has grown relative to enrollment at 4-yr institutions: In 1970, about 27% of all students enrolled in postsecondary education were at 2-yr institutions. (3) Beyond the secular trend in growing enrollment in 2-yr colleges, the recent economic downturn has been cited as an important factor behind large increases in enrollment in community colleges across the country. Because of open enrollment policies, low tuition, and their applied courses, community colleges can be relatively attractive during difficult economic times for adults looking to upgrade skills and traditional college students facing financial hardship. While national data on enrollments subsequent to the deteriorating economic conditions in late 2008 are not yet available, there is ample evidence of large increases in enrollment from 2-yr colleges across the country. (4)

As enrollment has continued to grow, we have come to learn more about the earnings effect of education at community colleges. But, like policy makers, researchers have focused less on enrollment at these institutions than on enrollment at 4-yr colleges and universities. So, while some work has been done to understand the effect of community colleges on employment and earnings of students, we have much to learn about the role of community colleges in shaping the economic futures of students.

Improving our understanding of the economic returns to community college education is important for several reasons. First, community colleges provide educational opportunities to students who are typically economically disadvantaged, and whose academic preparation is typically not as strong. These are the students most at risk of being left behind by ongoing changes in the labor market. Second, community colleges are a principal mechanism for upgrading skills of those already in the workforce because of their open admissions policies and flexible courses of study, which include degree programs, certificates, and nondegree courses. Finally, as others have argued, community college enrollment is likely to be particularly amenable to policy intervention because they enroll a larger share of students affected by state and federal financial aid (Rouse, 1994).

In this paper, I make use of data from the 2000 follow-up of the National Education Longitudinal Survey (NELS) to extend what is known about the value of education at community colleges. I examine the effects of enrollment in community colleges on students' subsequent earnings. My first objective is to estimate the earnings effects of varying levels of credits earned in community colleges, separately from any credentials earned while enrolled. I do so because community colleges are often used as a means for students to engage in study not necessarily leading to a degree or certificate. My second objective is to examine whether credits earned in courses that are principally occupational or vocational provide a different return than credit hours earned in academic courses.

Below, I briefly describe what is known about the returns to community college education. I then describe the NELS data and the methods employed to assess the value of enrollment and degrees from community colleges. Finally, I present my results and discuss their implications.

II. COMMUNITY COLLEGE EDUCATION AND ECONOMIC OUTCOMES

There has been a substantial amount of research over the past many decades to identify the effect of postsecondary education on earnings and employment. Consistent with the Mincerian model, the evidence supports the view that more schooling leads to higher earnings. Young people have responded to those incentives, enrolling in college after high school in increasing numbers. (5)

The bulk of the work on this topic has focused on the 4-yr level, despite the substantial level of enrollment in community colleges. While some early papers found that community college education appeared to have little or no positive effect on subsequent earnings (e.g., see Dougherty, 1987), subsequent research has uniformly found substantial positive earnings effects. Using the National Longitudinal Survey Class of 1972 data (NLS-72), Grubb (1993) found that students completing the associate degree earned significantly more that did their high school educated counterparts, while those who did not complete the degree fared no better. Similarly, Grubb (1997), using the Survey of Income and Program Participation data for the various years in the 1980s and in 1990, reported positive earnings effects of subbaccalaureate credentials, but very low benefits for those failing to earn a credential or degree.

Kane and Rouse (1995) used data from the NLS-72 and National Longitudinal Survey of the Labor Market Experience of Youth, 1979 (NLSY79) and found that students who attain a baccalaureate degree experience a 10-20% increase in earnings as compared to an associate degree holder. Additionally, Kane and Rouse found the economic returns to a community college degree to be roughly 15-25% higher than a high school diploma. They also examined the return to coursework not leading to a degree, at both community colleges and 4-yr institutions. They found similar returns to such coursework at both types of institutions, with 1-yr coursework associated with approximately 5-8% increase in earnings. Gill and Leigh (1997) found similar results using the NLSY79 data.

Grubb (2002) reviewed the research on the returns to subbaccalaureate degrees and course-work at community colleges not leading to degrees. He reported most estimates find that individuals who complete associate degrees earn about 20-30% more than high school graduates with estimates for men being somewhat lower, while estimates for women are somewhat higher. He also concluded that 1 yr of coursework (without completing a degree) at either a 2- or a 4-yr school increases earnings by about 5-10%. Returns to certificates are varied and often not measurable, likely because of small sample sizes and small effect sizes.

Marcotte et al. (2005) used the 2000 follow-up of the NELS to estimate earnings effects of community college education. They estimated that full-time enrollment in a community college increases earnings between 5 and 8% for each year enrolled, even if no degree was received. They found, further, that earning an associate degree increases earnings by about 15-30%. Finally, they found limited evidence that students enrolled in occupational areas earn more than those enrolled in academic programs.

While Marcotte et al. (2005) provide evidence that enrollment in community college has a positive earnings effect, that work is limited in an important respect. The authors made use of self-reported information on the timing and intensity of enrollment. Furthermore, they attempted to identify the effect of enrollment in occupational versus academic courses of study using self-reported information on first declared major. Using such self-reported information limits confidence in conclusions about the effect of enrollment not leading to degrees, and the effects of study in vocational or academic areas.

Both of these limitations may be important. Enrollment in a community college often does not lead to a degree. Using the Integrated Postsecondary Education Data System, Bailey et al. (2006) found that only between 10 and 30% of community college students earned degrees. Many students enroll in community colleges with the intent to learn a particular skill, not obtain a degree. So, much of the important variation in the "treatment" provided by a community college education is variation in courses, not degrees.

In this paper, I attempt to provide additional insight into the importance of course taking at community colleges for employment outcomes. I do so by using data from the postsecondary transcripts of the NELS cohort to identify the total number of credit hours completed by students. From these, I estimate the value of credit hours on earning and employment. I also determine the proportion of credit hours taken in vocational versus academic courses to examine whether taking vocational courses is especially advantageous. There has been a substantial focus on occupational education in community colleges over the past few decades, and the relative earnings of enrollment in courses with a vocational rather than academic focuses is of special concern (Alfonso, Bailey, and Scott, 2005).

III. DATA AND METHODS

The NELS follows a nationally representative sample of students who were in the eighth grade in 1988. The original survey included nearly 25,000 students. But, several thousand students were subsequently dropped due to budget constraints. Of those selected for follow up, interviews were conducted in 1990, 1992, 1994, and 2000. By 2000, data were collected from 11,559 of the original sample. In addition to the data collected on these students, the NELS also includes data collected from parents, teachers, and principals.

I restrict my analysis to NELS sample members who completed interviews in all years, and who were not still enrolled in a postsecondary institution in 1999 or at the time of the 2000 interview. I make this latter restriction to ensure that earnings were measured only for those respondents who had completed their education. I also eliminate those who reported earning a graduate or professional degree and those whose only postsecondary enrollment occurred after 1994. This was necessary because …