Download presentation
Presentation is loading. Please wait.
Published byImre Jenő Hegedüs Modified over 5 years ago
1
Students’ assessment of higher education in Spain
César Alonso-Borrego (Universidad Carlos III) Antonio Romero-Medina
2
Students' assessment of higher education in Spain
Main facts Huge increase in human capital since the 80s’. Currently, high % of young people with higher education. However, significant imbalances. Excess of long-degree university graduates. Lack of Higher Technical graduates. See San Segundo (1997, Cuadernos de ICE): Relatively low educational attainment in Spain before Currently, % of young college educated above OECD avg. and close to US. Lack of short degree graduates: San Segunmdo (2002, CRUE) Low % of non-university higher educated (necessary to accomplish skilled blue collar jobs) SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
3
Students' assessment of higher education in Spain
Motivation These imbalances may lower the relative return to college education. Aim: measure economic value of college education using student reported measures of Wages expected after graduation Shadow wages Such imbalances can make college education unattractive Measures of Wages expected after graduation in jobs that requires college education Shadow wages (wage that would make the student willing to early leave college) SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
4
Students' assessment of higher education in Spain
Main findings Expected wages. Major determinant: (recent) academic performance. Increase with the academic year. Shadow wages. Major determinant: academic performance and family background. Increase with academic year for males, steady for females. SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
5
Theoretical framework
Stylized model of human capital. Wage determinants Years of education and… Family background (human capital, income) Academic performance Other student characteristics (earlier choices, Gender?…) Under certain conditions the log of earnings can be expressed as a function of individual characteristics (family background, academic performance, academic choices). These covariates affect both objective and subjective predictions of student wages. Regarding subjective predictions, these are affected by the quality of student information (captured by these very same covariates). SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
6
Theoretical predictions
Covariates determine amount and quality of student information to compute wage expectations. Both expected and shadow wages should increase with academic year. Expected wage should be lower than shadow wage. Expected wage should be lower than shadow wage: Otherwise, student would be prompted to drop out early. SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
7
Students' assessment of higher education in Spain
Data Survey data about college student characteristics and attitudes toward college education. Info of student declarations about her expected and shadow wages (discretized). Three independent cross sections. 2001, 2004, 2005. Expected wage: What is the monthly wage that you are expecting after graduation? Shadow wage: What is the minimum monthly wage at which you would leave university before completion in exchange of an indefinite term contract with such real wage for your entire labor lifetime? SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
8
Students' assessment of higher education in Spain
Table 1. Monthly expected and shadow wages. Relative frecuencies (%) Expected Shadow Between 450 and 901 € 4.89 2.07 Between 901 and 1803 € 17.87 14.51 Between 1803 and 3606 € 52.66 17.94 Between 3606 and 5409 € 13.79 28.15 More than 5409 € 10.80 37.32 No. obs. 1371 1254 Kendall coef. ordinal correlation between expected and shadow wages: 0.28 (p-val: 0.01%) No response frequency: about 33%. Expected wages: unimodal (53% between 1803 and 3606 € per month). Shadow wages: distributed much more uniformly for any category above the minimum substantial level of right censoring (37% chose the upper category) SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
9
Students' assessment of higher education in Spain
Some differences by gender in: Family income Academic performance: pre-university (higher % of women passed access exam at 1st attempt) choices (higher % of women studying their 1st choice, lower % of women studying science) lower percentage of women repeating, women more satisfied with college. SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
10
Students' assessment of higher education in Spain
Empirical approach Baseline model However, W* unobserved. Instead, we observe Ordered response model with known thresholds SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
11
Empirical approach (ii)
Baseline model However, W* unobserved. Instead, we observe Ordered response model with known thresholds Thus, parameter scale is identified. If information on threshold values is not exploited, an additional assumption on the right tail of the wage distribution (for declared expected monthly wages above 5409 €) is needed. In such a case, predictions on individual expected wages are very sensitive to this additional assumption. SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
12
Empirical approach (iii)
No evidence that atrittion due to non response about wages produced bias. Two subsamples according with academic year: First Year Penultimate Year Interactions with type of studies Subsamples: 1st degree year, and the year before the last degree year. We would expect the effects of the conditioning variables to differ very much for these two particular groups (extreme cases according with the time horizon for graduation). Namely, we would expect students nearer to completion to have much lower uncertainty about their academic prospects better knowledge of their job market prospects after graduation. Obviously, there exist substantial differences among the distribution of these two subsamples (adverse selection of first year students in terms of quality and probability of becoming graduates). SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
13
Students' assessment of higher education in Spain
Reference group: Male at non science short degree SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
14
Students' assessment of higher education in Spain
SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
15
Results: expected wages
Lower for female, esp. when nearer to completion. Relevance of nearer academic performance. Family background: positive effect. Female effect turns positive for later year female sudents in science disciplines. Relevance of nearer academic performance. Pre-university variables (1st year) University performance –Repeat, Grant, Satisfied- (later year). Overestimation of expected wages with respect to actual wages for working graduates aged is actually larger than the difference shown, because individuals in our sample are not strictly comparable with the sample from which average actual wages has been computed, which restricts to graduates aged 20 to 29 who have decided to work and have found a job. Our student sample is composed of individuals who are not graduated. For those who will become graduates, a percentage of them will eventually not work, either because they decide not to enter the labor market, or because they will not find a job. Besides, usually an adverse selection of them will drop out college before completion. Therefore, part of the apparent improvement in expectation formation as degree years increase merely reflects a sample selection of students who are much more likely to work in jobs that require university education. SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
16
Results: expected wages (ii)
Predictions More realistic for later year students. Lower for female. Gaps of prediction w.r.t. actual wages Lower for later year and long degree student Female effect turns positive for later year female sudents in science disciplines. Relevance of nearer academic performance. Pre-university variables (1st year) University performance –Repeat, Grant, Satisfied- (later year). Overestimation of expected wages with respect to actual wages for working graduates aged is actually larger than the difference shown, because individuals in our sample are not strictly comparable with the sample from which average actual wages has been computed, which restricts to graduates aged 20 to 29 who have decided to work and have found a job. Our student sample is composed of individuals who are not graduated. For those who will become graduates, a percentage of them will eventually not work, either because they decide not to enter the labor market, or because they will not find a job. Besides, usually an adverse selection of them will drop out college before completion. Therefore, part of the apparent improvement in expectation formation as degree years increase merely reflects a sample selection of students who are much more likely to work in jobs that require university education. SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
17
Students' assessment of higher education in Spain
SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
18
Students' assessment of higher education in Spain
SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
19
Students' assessment of higher education in Spain
Results: shadow wages Gender effect small. Relevance of family background: university father (+) Family income (+) Relevance of nearer academic performance. Importance of reasons behind degree choice. Predictions More realistic for later year students. Lower for female. Gaps of prediction w.r.t. actual wages Much higher for female. Gender effect small: so, differences in predictions will arise because of differences among men and women, and not in the mere fact of being either men or women. Relevance of family background variables decreases for students in science disciplines. The fact that relative shadow wages are much higher for females, even though they are worse paid than males, is in accordance with the fact that women enjoy a higher relative differencial return to university education w.r.t. other educational levels. (Investment in college education more attractive for men than for women). SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
20
Students' assessment of higher education in Spain
SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
21
Students' assessment of higher education in Spain
Drop-out propensity Drop-outs not observed in our sample. Evidence on large college drop-out rate and narrow wage distribution in Spain. But those declaring shadow wage below her expected wage are more prompted to leave. SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
22
Students' assessment of higher education in Spain
SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
23
Students' assessment of higher education in Spain
SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
24
Results: potential drop out
Potential drop-out characteristics: Poor academic performance Part-time student Low family human capital A student who declares a shadow wage lower than her expected wage believes that her wage profile along her working life as a graduate will not compensate the cost of finishing her studies. There are policy implications derived from these findings. The wage distribution by education level in Spain is relatively narrow, so that the return to higher education is small relative to other OECD countries. In fact, the drop out rate in Spain is remarkably large, what is mostly attributed to the failure of the educational system. Our analysis points out an alternative explanation. There are economic reasons, determined by observed variables, that can explain, at least partially, the drop out propensity found in the Spanish university. SAE 2008, Zaragoza December 11, 2008 Students' assessment of higher education in Spain César Alonso-Borrego
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.