THE LABOR MARKET RETURN IN EARNINGS TO COMMUNITY COLLEGE CREDITS AND CREDENTIALS PETER RILEY BAHR, ASSOCIATE PROFESSOR

THE LABOR MARKET RETURN IN EARNINGS TO COMMUNITY COLLEGE CREDITS AND CREDENTIALS PETER RILEY BAHR, ASSOCIATE PROFESSOR CENTER FOR THE STUDY OF HIGHER AND POSTSECONDARY EDUCATION

BEFORE CITING THIS STUDY, PLEASE CONTACT THE AUTHOR (PETER RILEY BAHR; PRBAHR@UMICH.EDU) FOR THE FULL PAPER ON WHICH IT IS BASED Slide 2

VERY PLEASED TO BE PRESENTING TODAY This research is brand new, and you are the first audience to hear the results. As a point of clarification, this research is an extension of the missing gpieces research on skills-builders but it addresses a MUCH broader population of nearly all first-time students. It was a big project (94 pages at last count) and I will focus on just one slice of the research today. Please feel free to interrupt with questions. Slide 3

AN UNCONVENTIONAL BACKGROUND Attended Solano Community College Four associate degrees, including one in water & wastewater tech Worked in a wastewater treatment plant for eight years while completing my undergraduate education BA in Criminology from CSU Sacramento Worked as researcher in the Chancellor s Office while attending graduate school MA and PhD in Sociology from UC Davis Former Assistant Professor of Sociology at Wayne State U. Now serving as Associate Professor of Higher Education at the University of Michigan Slide 4

WHAT DO WE MEAN BY LABOR MARKET RETURNS? Slide 5

WHAT ARE LABOR MARKET RETURNS? Typically refer to: Earnings (per quarter or per year) Hourly Wages Employment (employed versus unemployed) Here, I am focusing on earnings. I am working currently on a study of employment as well. Slide 6

WHY IS STUDYING STUDENTS EARNINGS DIFFICULT? Slide 7

FIVE SIGNIFICANT CHALLENGES The inflation problem The missing sectors problem The high school earnings problem The frozen yogurt shop problem The unobserved heterogeneity problem Slide 8

THE INFLATION PROBLEM Inflation drives up earnings, regardless of educational attainment. Fairly easy to solve with inflation adjustment which makes the buying power of dollars in past years comparable to the buying power of today s dollars. Slide 9

THE MISSING SECTORS PROBLEM Different sources of information about earnings cover (observe) various employment sectors to differing degrees. Unemployment Insurance (UI) data do not address: self-employment l military employment and federal civilian employment employment through informal cash arrangements employment in other states To the extent that a field of study leads to employment in one of these sectors, UI data will underestimate earnings. Slide 10

THE HIGH SCHOOL EARNINGS PROBLEM Measuring the return in earnings to postsecondary education typically involves pre-college (or pre-credential) and post- college measurements. If the pre-college (or pre-credential) measures include a period of time in which students were in high school relative earnings in the post-college period will be inflated artificially. The same holds for pre-credential measures of labor market returns collected while a student was attending college. Slide 11

THE FROZEN YOGURT SHOP PROBLEM Given enough time, the newly-hired, part-time employee at the frozen yogurt shop will eventually advance to full-time assistant manager with or without a postsecondary education. Even after removing the effect of inflation, an individual s earnings tend to rise over time due to increasing i experience. To measure the effect of education on earnings, one must account for the underlying earnings trend. Slide 12

THE UNOBSERVED HETEROGENEITY PROBLEM Unmeasured differences between individuals that are systematically related to earnings and educational attainment. For example Program of study is related to earnings. Sex is related to program of study. Sex is independently related to earnings: women are more likely to have interrupted employment histories and, therefore, lower average earnings. Some of the relationship between program of study and earnings is a result of differences between males and females in patterns of choice. Slide 13

WHY TALK ABOUT THESE CHALLENGES? Different approaches to studying the relationship between education and earnings can produce very different results in large part because they address these challenges in different ways and to different degrees. My study uses an approach that is different from Salary Surfer and probably different from the approaches that you have used at your college. Comparing results requires sensitivity to differences in the goals of the research and the underlying methodology. Seemingly different results in fact may be similar once differences in analytical conditions are taken into account. Slide 14

WHAT DOES PRIOR RESEARCH TELL US ABOUT THE RETURNS TO NON COMPLETING STUDENTS? Slide 15

RETURNS TO NON COMPLETING STUDENTS Older research generally treated non-completing students as if every credit is of equal economic value or divided credits into two very broad categories, such as vocational versus academic. Recent research has neglected the labor market returns experienced by non-completing students, t focusing almost exclusively on the returns to credentials. Recent research has treated non-completing students as a homogenous and un-quantified comparison group against which h the returns to credentials are measured. Slide 16

RETURNS TO NON COMPLETING STUDENTS All prior research has neglected to carefully disentangle the human capital acquired in coursework from the signaling value of the credentials. As a result, we know very little about how non-completing students fair in the labor market. Yet, more than half of community college students do not complete a credential or transfer to a four-year institution. Moreover, research indicates that non-completing students often pursue coherent course-taking pathways in CTE fields. Slide 17

WHAT DOES THE RESEARCH TELL US? Slide 18

WHAT DOES THE RESEARCH TELL US? Broadly speaking (and with the caveat noted previously) Associate degrees and long-term certificates (requiring more than one year to complete) generally result in decent earnings gains especially those in CTE fields. Short-term certificates (those requiring a year or less to complete) generally are not beneficial to students earnings. CTE credits (treated as an undifferentiated mass) without a credential tend to be of modest value. Slide 19

MY STUDY Slide 20

MY SUPPOSITION If the returns to a CC education are mainly a result of the human capital acquired in the coursework then the earnings gains experienced by students should correspond to the coursework that they take and the credits that they complete. Consequently, the return to non-completing students could rival that of completing students. If the return to a CC education is mainly a result of signaling then the earnings gains experienced by non-completing students should be very small. Slide 21

QUESTIONS TO ANSWER Ignoring the distinction between human capital and signaling, how do the returns to credentials in California differ from those observed other states? Are the observed effects of credentials on earnings strictly a consequence of the credential itself (e.g., signaling) and therefore something that only completing students experience? Or, are they a product of the skills and knowledge (human capital) that were acquired in the coursework and therefore something that non-completing students also experience? Slide 22

ANALYTICAL SAMPLE 759,489 first-time students with valid SSNs Between 18 and 50 years old at college entry Entered the CCC system between Fall 02 and Summer 06 At least one UI quarterly earnings record before entering the CCC system, and at least one after entering the CCC system Tracked UI quarterly earnings from 10 quarters prior to entering the CCC system through the fourth quarter of 2012 9.00 to 12.75 years, depending on term/year of CCC entry Confined analysis to quarters in which students had non-zero earnings Total of 22,215,020 student-quarters included in the analysis Slide 23

METHOD OF ANALYSIS Adjusted earnings for inflation Student-level fixed-effects regression model Controls for all unobserved, time-invariant differences between students, such as motivation, intelligence, etc. Addresses much of the unobserved heterogeneity problem Included d a complex earnings trend Allowed to vary by sex, race/ethnicity, age at college entry, citizenship, academic goal, whether a student ever transferred to a four-year institution, and term/year of college entry Addresses much of the frozen yogurt shop problem Slide 24

METHOD OF ANALYSIS Other statistical controls: Credits attempted (quadratic) in each term Grant aid in each term Loan aid in each term Enrollment in non-ccc postsecondary institutions in each term Focal variables Four levels of credentials in 23 fields of study Cumulative completed credits (quadratic) in 181 subfields Slide 25

ESTIMATED EARNINGS GAINS FOR CREDENTIALS RELATIVE TO NON COMPLETING STUDENTS (NO PARSING OF NON COMPLETING PATHWAYS) Slide 26

AVERAGE PERCENTAGE CHANGE IN EARNINGS Low Credit Short Term Long Term Assoc Award Certificate Certificate Degree Field of Study of Credential < 6 units 6 29 units 30+ units 60+ units Health +8% +11% +39% +106% Business & Management +12% +9% N/S +6% Public & Protective Services +13% +32% +27% +11% Engineering & Industrial Tech N/S +11% +11% +12% Family & Consumer Sciences N/S +9% +6% +3% Commercial Services ---------- -6% -16% -14% Information Technology ---------- +9% N/S +12% Slide 27

AVERAGE PERCENTAGE CHANGE IN EARNINGS Low Credit Short Term Long Term Assoc Award Certificate Certificate Degree Field of Study of Credential < 6 units 6 29 units 30+ units 60+ units Media & Communications ---------- N/S N/S -8% Agriculture & Natural Resources N/S N/S N/S +16% Law ---------- N/S +14% +12% Architecture & Related Tech N/S N/S N/S N/S Environmental Sciences & Tech N/S ---------- ---------- N/S Library Science ---------- +27% ---------- N/S Slide 29

AVERAGE PERCENTAGE CHANGE IN EARNINGS Low Credit Short Term Long Term Assoc Award Certificate Certificate Degree Field of Study of Credential < 6 units 6 29 units 30+ units 60+ units Media & Communications ---------- N/S N/S -8% Agriculture & Natural Resources N/S N/S N/S +16% Law ---------- N/S +14% +12% Architecture & Related Tech N/S N/S N/S N/S Environmental Sciences & Tech N/S ---------- ---------- N/S Library Science ---------- +27% ---------- N/S Slide 30

TAKE AWAYS The earnings gains associated with CTE credentials are largely positive and oftentimes strong. In contrast to claims that short-term certificates have no labor market value earnings gains hold nearly as well for short-term awards as they do for long-term awards. Even some of the low-credit awards (< 6 units) show positive earnings gains. Slide 31

ESTIMATED EARNINGS GAINS FOR CREDENTIALS AFTER PARSING THE VALUE OF CREDENTIALS FROM THE VALUE OF CREDITS IN 181 SUBFIELDS Slide 32

BEFORE PARSING OUT THE VALUE OF HUMAN CAPITAL Low Credit Short Term Long Term Assoc Award Certificate Certificate Degree Field of Study of Credential < 6 units 6 29 units 30+ units 60+ units Health +8% +11% +39% +106% Business & Management +12% +9% N/S +6% Public & Protective Services +13% +32% +27% +11% Engineering & Industrial Tech N/S +11% +11% +12% Family & Consumer Sciences N/S +9% +6% +3% Commercial Services ---------- -6% -16% -14% Information Technology ---------- +9% N/S +12% Slide 33

AFTER PARSING OUT THE VALUE OF HUMAN CAPITAL Low Credit Short Long Assoc Award Certificate Certificate Degree Field of Study of Credential < 6 units 6 29 units 30+ units 60+ units Health +6% +10% +26% +99% Business & Management +10% +5% N/S N/S Public & Protective Services N/S +13% +10% N/S Engineering & Industrial Tech N/S N/S N/S +7% Family & Consumer Sciences N/S +5% N/S N/S Commercial Services ---------- +9% N/S N/S Information Technology ---------- N/S N/S N/S Slide 34

BEFORE PARSING OUT THE VALUE OF HUMAN CAPITAL Low Credit Short Term Long Term Assoc Award Certificate Certificate Degree Field of Study of Credential < 6 units 6 29 units 30+ units 60+ units Media & Communications ---------- N/S N/S -8% Agriculture & Natural Resources N/S N/S N/S +16% Law ---------- N/S +14% +12% Architecture & Related Tech N/S N/S N/S N/S Environmental Sciences & Tech N/S ---------- ---------- N/S Library Science ---------- +27% ---------- N/S Slide 36

AFTER PARSING OUT THE VALUE OF HUMAN CAPITAL Low Credit Short Long Assoc Award Certificate Certificate Degree Field of Study of Credential < 6 units 6 29 units 30+ units 60+ units Media & Communications ---------- N/S N/S N/S Agriculture & Natural Resources N/S N/S N/S +11% Law ---------- N/S N/S N/S Architecture & Related Tech N/S N/S N/S N/S Environmental Sciences & Tech N/S ---------- ---------- N/S Library Science ---------- N/S ---------- N/S Slide 37

TAKE AWAYS The earnings gains associated with CTE credentials that are independent of course-taking pathways generally are much smaller or, more often, zero. Does this mean that CTE credentials have little or no value? No, it means that the value of the credentials is driven by the human capital (the skills and knowledge) that students acquire. So, what is the value of that human capital? Slide 39

ESTIMATED EARNINGS GAINS FOR COMPLETED CREDITS BY SUBFIELD OF STUDY TOP 30 MOST POPULAR CTE SUBFIELDS (82.4% OF ALL CTE CREDITS) Slide 40

AVERAGE PERCENTAGE CHANGE IN EARNINGS Subfield of Study of Credits % of All CTE Units 6 Units 12 Units 18 Units Early Childhood Education 10.00% +1.5% +3.3% +5.5% Administration of Justice 8.86% +6.8% +13.6% +20.3% Nursing 5.93% -2.5% -3.9% -4.1% Accounting 5.74% +4.4% +8.0% +11.0% Construction Crafts Technology 4.08% +3.2% +6.2% +9.0% Cosmetology & Barbering 3.62% -5.9% -10.9% -14.8% Fire Technology 3.56% +3.8% +7.9% +12.4% Office Technology 3.56% +2.5% +4.5% +6.0% Nutrition, Foods & Culinary Arts 3.49% -1.1% -1.8% -1.9% Business & Commerce, General 3.47% +1.2% +1.5% +1.1% Slide 41

AVERAGE PERCENTAGE CHANGE IN EARNINGS Subfield of Study of Credits % of All CTE Units 6 Units 12 Units 18 Units Automotive Technology 3.46% +1.6% +3.2% +4.7% Information Technology, General 3.31% +2.8% +5.0% +6.6% Business Management 2.63% +0.7% -0.5% -3.7% Computer Information Systems 2.23% N/S N/S N/S Manufacturing & Industrial Tech 2.00% +5.5% 5% +10.7% +15.4% Real Estate 1.71% -5.5% -8.9% -10.6% Emergency Medical Services 1.51% -0.4% +0.4% +2.6% Business Administration 1.43% +1.2% -0.7% -5.6% Human Services 1.39% N/S N/S N/S Electronics & Electric Technology 1.39% +6.7% +12.6% +17.5% Slide 43

AVERAGE PERCENTAGE CHANGE IN EARNINGS Subfield of Study of Credits % of All CTE Units 6 Units 12 Units 18 Units Digital Media 1.16% N/S N/S N/S Computer Software Development 1.08% +4.6% +8.4% +11.4% Drafting Technology 0.99% N/S N/S N/S Radio & Television 0.92% -4.1% -7.1% -9.2% Applied Photography 0.88% -4.4% 4% -9.2% -14.4% 4% Fashion 0.82% N/S N/S N/S Dental Occupations 0.80% -2.0% -3.5% -4.5% Computer Infrastructure & Support 0.80% +4.6% +9.0% +13.2% Marketing & Distribution 0.78% N/S N/S N/S Radiologic Technology 0.75% -8.6% -14.8% -18.8% Slide 45

TAKE AWAYS The majority of CTE credits are associated with significant and frequently strong earnings gains... independent of awarded credentials. Earnings gains in several fields are especially consistent: engineering & industrial technologies public & protective ti services information technology Non-completing students can experience earnings gains that are as large, or larger than, those experienced by completing students, t depending di on course-taking pathway. Slide 47

CONCLUSION Earnings gains can be found at all levels of educational attainment in the CCC system including completing credits but not a credential. The current nationwide focus on improving rates of completion is unquestionably good. However, when this focus is enacted in policies that limit access for students who are pursuing non-credential goals unintended negative consequences will result. Slide 48

PETER RILEY BAHR PRBAHR@UMICH.EDU