Abstract
This paper analyzes the relative importance of short-term financial constraints vis-a-vis skills and other background factors when explaining higher education access in Peru. We focus on the disparities in university enrollment between rich and poor households. We use a novel household survey that includes special tests to measure cognitive and socio-emotional skills of the urban population age 14–50. These are complemented with retrospective data on basic education and family socioeconomic conditions in a multinomial model. We find that the strong correlation between university enrollment and family income in urban Peru is not only explained by short-term credit constraints, but also by poor cognitive skills and by family and educational backgrounds affecting tastes and aptitudes for formal education. Family income explains, at most, half of the university access gap between poor and non-poor households. The other half is related to differences in parental education, educational backgrounds, and cognitive skills. Our results indicate that credit or scholarship schemes alone will not suffice to change the regressive nature of higher education enrollment in Peru, and that such programs will face strong equity–efficiency trade-offs.
This is a preview of subscription content, log in via an institution to check access.
Source: 2010 national household survey (ENAHO 2010)
Source: Asamblea Nacional de Rectores
Source: ENHAB (2010)
Source: ENHAB (2010)
Source: ENHAB (2010)
Similar content being viewed by others
Notes
See Holm-Nielsen et al. (2005) for a comparison across Latin American countries.
It is well documented for Peru that average returns for university higher education are large (around 17 % after factoring in direct costs) and considerably larger than returns for technical higher education (see Yamada and Castro 2010).
Some high-quality private universities rely on interviews or essay-based examinations for applicants with outstanding academic performance during secondary education. These forms of entry have limited reach as they apply only for students graduating from elite private schools.
A solid body of evidence from biology (epigenetics), neuroscience, psychology, and education supports a consensus that relying solely on “Nature” or “Nurture” to explain later-life outcomes is obsolete. This evidence vindicates the power of public intervention to influence cognitive and socio-emotional skills (Shonkoff and Phillips 2000; Cunha and Heckman 2008).
We decided to group these last two categories because only 0.63 % of the sample reported a high socioeconomic status. Proportions of each category in the sample are: 59.2 % low and 40.8 % medium–high.
This latter effect can be particularly important when the educational system exhibits significant heterogeneity in terms of quality, like in Peru.
Contributions are based on a multinomial logit specification. The mathematical expressions for these contributions are available from the authors upon request.
This is explanation provided by Heckman and Rubinstein (2001) for the wage differentials between high school graduates and GED recipients despite having equivalent credentials and measured cognitive skills.
Absence of correlation between error terms confirms that there is no risk of selection bias arising from high school dropouts. Coefficient signs and significance in the bivariate probit model for the probability of higher education enrollment given high school completion are consistent with our main multinomial model estimates.
The mathematical derivation of these biases is available from the authors upon request.
Instrumental variable estimation could be applied to obtain a consistent estimate of the effect of education on measured skills. This estimate could then be used to remove the effect of educational attainment and work with an adjusted version of test scores. The success of this strategy heavily depends on the choice of instrument: it should correlate with educational attainment but should not correlate with the skills. This is difficult to accomplish because skills are the result of a cumulative process and the quantity and quality of their early childhood determinants typically correlate with educational attainment.
We tried several versions of an adjusted test score supposedly clean of the effect of higher education attendance. Instruments used to obtain a consistent estimate of the effect of educational attainment on test scores included distance to school and age, the latter based on the notion that if age affects measured skills in a sample of high school graduates, it should only be through granting more opportunities to attend higher education. In both cases the results obtained suggested we were underestimating the effect of skills on higher education enrollment: estimated marginal effects were not significant or even negative.
Note that in panel (A) we already control for individual characteristics such as age, sex, and language.
Concerns can arise regarding the use of retrospective data to reflect the availability of monetary resources by the time post-secondary choices were made. For example, more educated people enjoying a better economic situation at present can tend (by comparison) to understate their past socioeconomic condition. This situation would bias downward our estimate of the effect of family income on higher education access. We tested this possibility by analyzing whether, for a given background, the percentage of respondents who recalled a “low” socioeconomic status raised with higher education access. As expected, in the general population, the percentage of respondents who recall a “low” socioeconomic condition falls from 73 to 48 % between those who did not access higher education and those who did have access. A similar pattern is observed when we fix individuals’ background. In the sample of individuals who attended a public school outside Lima and whose parents have only up to secondary education, the percentage of respondents who recall a “low” socioeconomic condition falls from 76 to 53 % with higher education access. Thus, there is no indication that access to higher education leads to individuals reporting lower socioeconomic conditions present at the time post-secondary choices were being made.
Educational background variables could also be capturing the heterogeneous effect of skills in different skill groups. We added quadratic terms for skills and dummies allowing for different effects depending on the position in the skill distribution. None of these were significant, and the results shown in panel D were robust to these specifications.
In Peru, children belonging to more affluent families have access to a more nurturing early childhood environment and also enjoy better preschool and basic education opportunities (Cueto et al. 2014).
The number of public universities also grew, but in a much smaller proportion. They went from 28 to 35 between 1996 and 2010.
Peru ranked 63 and 64 out of 65 countries in the 2009 PISA evaluations in reading comprehension and mathematics, respectively. Only 13 % of secondary students assessed were able to pass the level 3 threshold in reading comprehension (moderate complexity) against an average of 60 % in OECD countries.
References
Barrick, M. R., & Mount, M. K. (2005). Yes, personality matters: Moving on to more important matters. Human Performance, 18(4), 359–372.
Barrick, M. R., Mount, M. K., & Judge, T. A. (2001). Personality and performance at the beginning of the new millennium: What do we know and where do we go next? International Journal of Selection and Assessment, 9(1–2), 9–30.
Bassi, M., & Galiani, S. (2010). Labor market insertion of young adults in Chile”. Mimeo. Inter-American Development Bank.
Becker, G. S. (1964). Human capital: A theoretical and empirical analysis, with special reference to education. New York: National Bureau of Economic Research, distributed by Columbia University Press.
Bowles, S., & Gintis, H. (1976). Schooling in capitalist America: Educational reform the contradictions of economic life. New York: Basic Books.
Castro J. F., & Yamada, G. (2012). “Convexification” and “Deconvexification” of the Peruvian wage profile: A tale of declining education quality. Universidad del Pacifico working paper DD/12/02.
Cameron, S. V., & Heckman, J. J. (2001). The dynamics of educational attainment for black, Hispanic, and white males. Journal of Political Economy, 109(3), 455–499.
Card, D. (1994). Earnings, schooling and ability revisited. National Bureau of economic research working paper no. 4832.
Card, D., & Krueger, A. (1996). Labor market effects of school quality: Theory and evidence. National Bureau of economic research working paper no. 5450.
Carneiro, P., & Heckman, J. J. (2002). The evidence on credit constraints in post-secondary schooling. Economic Journal, 112(482), 705–734.
Castro, J. F., & Rolleston, C. (2015). Explaining the urban–rural gap in cognitive achievement in Peru: The role of early childhood and school influences. Young lives working paper no. 139.
Castro, J. F., & Yamada, G. (2010). Educación superior e ingresos laborales: estimaciones paramétricas y no paramétricas de la rentabilidad por niveles y carreras en el Perú. DD/10/06, Universidad del Pacifico.
Castro, J. F., & Yamada, G. (2013). “Declining higher education quality affects postsecondary choices: A Peruvian case. International Higher Education, No. 70, Winter. pp. 26–28.
Checchi, D. (2006). The economics of education. New York: Cambridge University Press.
Claux, M., & La Rosa, M. (2010). Estudio de factores relacionados con la empleabilidad en zonas urbanas del Perú. Desarrollo de escalas de personalidad y emprendimiento. Unpublished manuscript.
Cueto, S., Guerrero, G., Leon, J., Zapata, M., & Freire, S. (2014). The relationship between socioeconomic status at age one, opportunities to learn and achievement in mathematics in fourth grade in Peru. Oxford Review of Education, 40, 50–72.
Cueto, S., Muñoz, I., & Baertl, A. (2010). Scholastic achievement, cognitive skills and personality traits of youths and adults in Peru: A cross-sectional and intergenerational analysis. Unpublished manuscript. GRADE, Lima.
Cunha, F., & Heckman, J. (2008). Formulating, identifying and estimating the technology of cognitive and noncognitive skill formation. Journal of Human Resources, 43, 738–782.
Cunha, F., Heckman, J., Lochner, L., & Masterov, D. (2006). Interpreting the evidence on life cycle skill formation. In E. A. Hanushek & F. Welch (Eds.), Handbook of the economics of education, Chapter 12 (pp. 697–812). Amsterdam: North-Holland.
Diaz, J. J., Arias, O., & Tudela, D. V. (2013). Does perseverance pay as much as being smart? The returns to cognitive and non-cognitive skills in urban Peru. Working paper. GRADE.
Duckworth, A. L., Peterson, C., Matthews, M. D., & Kelly, D. R. (2007). Grit: Perseverance and passion for long-term goals. Personality Processes and Individual Differences, 92(6), 1087.
Eccles, J., & Wigfield, A. (2000). Schooling’s influences on motivation and achievement. In S. Danziger & J. Waldfogel (Eds.), Securing the future: Investing in children from birth to college. New York: Russell Sage Foundation.
Farkas, G. (2003). Cognitive skills and noncognitive traits and behaviors in stratification processes. Annual Review of Sociology, 29, 541–562.
Frempong, G., Ma, X., & Mensah, J. (2012). Access to postsecondary education: Can schools compensate for socioeconomic disadvantage? Higher Education, 63(1), 19–32.
Garfias, M. (2015). La persistencia de las desigualdades en el ámbito de la educación universitaria. El caso de la Universidad Nacional Mayor de San Marcos, 1940–2000. In R. Cuenca (Ed.), La educación universitaria en el Perú: democracia, expansión y desigualdades. Serie Estudios sobre Desigualdad (10th ed.). IEP: Lima.
Goldberg, L. R. (1990). An alternative “description of personality”: The big-five factor structure. Journal of Personality and Social Psychology, 59(6), 1216–1229.
Hansen, K. T., Heckman, J. J., & Mullen, K. (2004). The effect of schooling and ability on achievement test scores. Journal of Econometrics, 121, 39–98.
Heckman, J. J. (2000). Policies to foster human capital. Research in Economics, 54(1), 3–56. (With discussion).
Heckman, J. J., & Rubinstein, Y. (2001). The importance of noncognitive skills: Lessons from the GED testing program. American Economic Review, 91(2), 145–149.
Heckman, J. J., Savelyev, P., Yavitz, A. (2006a). The Perry preschool project: A reanalysis. Unpublished manuscript. University of Chicago, Department of Economics.
Heckman, J., Stixrud, J., & Urzua, S. (2006b). The effects of cognitive and noncognitive abilities on labor market outcomes and social behavior. Journal of Labor Economics, 24(3), 411–482.
Heckman, J. J. (2006c). Skill formation and the economics of investing in disadvantaged children. Science, 312, 1900–1902.
Holm-Nielsen, L., Thorn, K., Brunner, J., & Balan, J. (2005). Regional and international challenges to higher education in Latin America. In H. De Wit, C. Jaramillo, J. Gavel-Avila, & J. Knight (Eds.), Higher education in Latin America: The international dimension. Washington: The World Bank.
Jerrim, J., & Vignoles, A. (2015). University access for disadvantaged children: A comparison across countries. Higher Education, 70(6), 909–921.
Lavado, P., Martínez, J. J., & Yamada, G. (2014). Una promesa incumplida? La calidad de la educación superior universitaria y el subempleo profesional en el Perú. BCRP working paper series no. 21- 2014. Lima, Banco Central de Reserva del Perú.
Morón, E., Castro, J. F., & Sanborn, C. (2009). Helping reforms deliver inclusive growth in Peru. In L. Rojas-Suarez (Ed.), Growing pains in Latin America. Washington: Center for Global Development (CGD).
Roberts, B. W., Kuncel, N. R., Shiner, R., Caspi, A., & Goldberg, L. R. (2007). The power of personality: The comparative validity of personality traits, socioeconomic status, and cognitive ability for predicting important life outcomes. Perspectives on Psychological Science, 2(4), 313–345.
Shonkoff, J. P., & Phillips, D. (2000). From neurons to neighborhoods: The science of early child development. Washington, DC: National Academy Press.
Spence, A. M. (1973). Job market signaling. Quarterly Journal of Economics, 87(3), 355–374.
Vargas, J. (2015). Navegando en aguas procelosas. Una mirada al sistema universitario peruano. In R. Cuenca (Ed.), La educación universitaria en el Perú: democracia, expansión y desigualdades. Serie Estudios sobre Desigualdad (10th ed.). IEP: Lima.
Wilson-Strydom, M. (2015). University access and theories of social justice: Contributions of the capabilities approach. Higher Education, 69(1), 143–155.
Acknowledgments
Authors would like to thank Roberto Asmat, Fernando Mendo, and David Vera-Tudela for their skillful research assistance and are thankful for useful insights and comments by participants at seminars at the World Bank in presentations of earlier versions of the paper. Juan Francisco Castro and Gustavo Yamada acknowledge financial support from the World Bank. The opinions expressed herein do not represent the views of the World Bank, its Executive Directors, or the governments they represent. All remaining errors are the authors’ responsibility.
Author information
Authors and Affiliations
Corresponding author
Appendices
Appendix 1: Encuesta nacional de habilidades (ENHAB) de Peru Urbano, 2010
Skills measurement
-
Sample: age 14–50, one randomly chosen (prefield) member per HH (n = 2666) without replacement (exclude illiterate, non-spanish speaker).
-
Cognitive tests (after pilot validation/revisions):
-
PPVT 4 (verbal perceptive skill, images are shown and must be matched to words, standardized protocol).
-
Verbal fluency (# valid P-words in 3 min).
-
Short-term memory (capacity to recall progressive sequence of digits read to test taker).
-
Numeracy/problem solving (18-item multiple choice test, timed 15 min).
-
Personality tests.
-
BFF 35-item bipolar adjectives, short-sentenced inventory (pretested in Lima student population), and 17-item grit scale (adapted to Peruvian context).
-
Special, intensified training and evaluation of enumerators (chose best).
-
US$10 incentive to participate. Applied in regular home environment though enumerators instructed to secure quiet space. Recorded data on administration conditions (time, duration, distraction, examiner FE).
-
Appendix 2
See Tables 3.
(A) Multinomial logit estimates for the complete model (“Did not enroll” = baseline category)
Covariate groups | Technical | University |
|---|---|---|
Socioeconomic status | ||
Medium or high = 1 | 0.41** | 0.869*** |
Cognitive skills | ||
Aggregate measure | 0.557*** | 1.02*** |
Socio-emotional skills | ||
Grit | 0.05 | 0.445*** |
Extraversion | 0.00 | −0.264* |
Agreeableness (“easy going”) | 0.12 | 0.07 |
Agreeableness (“reliable”) | −0.242** | −0.02 |
Conscientiousness | 0.02 | −0.17 |
Emotional stability | 0.08 | 0.03 |
Openness | 0.14 | −0.01 |
Parental background | ||
Father educational attainment (secondary = 1) | 0.604** | 0.558** |
Father educational attainment (higher = 1) | 0.676** | 1.326*** |
Mother educational attainment (secondary = 1) | −0.16 | −0.28 |
Mother educational attainment (higher = 1) | 0.06 | 0.10 |
Importance given by parents to education (high = 1) | 0.03 | −0.19 |
Importance given by mother to education (high = 1) | −0.52 | 0.682** |
Educational background | ||
Preschool (public = 1) | −0.511** | −0.373* |
Preschool (private = 1) | −0.18 | −0.76 |
Public school = 1 | 0.07 | −1.096*** |
Had to repeat a year or more in school = 1 | −0.27 | −0.76*** |
Perception regarding performance (top student = 1) | 0.45 | 0.984*** |
Perception regarding effort (large = 1) | 0.497*** | 0.525** |
Constant | −1.017* | −0.866° |
(B) Bivariate probit estimates used to test for selection within the sample of high school graduates
Covariate groups | Pr (complete high school) | Pr (enrolled in higher education|complete high school) |
|---|---|---|
Socioeconomic status | ||
Medium or high = 1 | 0.50*** | 0.27** |
Cognitive skills | ||
Aggregate measure | 0.56*** | 0.41*** |
Socio-emotional skills | ||
Grit | −0.05 | 0.11* |
Extraversion | 0.08 | −0.07 |
Agreeableness (“easy going”) | 0.02 | 0.03 |
Agreeableness (“reliable”) | −0.11 | −0.05 |
Conscientiousness | 0.04 | −0.04 |
Emotional stability | 0.01 | 0.01 |
Openness | 0.15* | 0.04 |
Parental background | ||
Father educational attainment (secondary = 1) | 0.34** | 0.32** |
Father educational attainment (higher = 1) | 0.69** | 0.62*** |
Mother educational attainment (secondary = 1) | 0.00 | −0.11 |
Mother educational attainment (higher = 1) | 0.06 | 0.03 |
Importance given by parents to education (high = 1) | 0.27** | −0.12 |
Importance given by mother to education (high = 1) | 0.08 | 0.09 |
Educational background | ||
Preschool (public = 1) | −0.30* | −0.29** |
Preschool (private = 1) | 0.12 | −0.29 |
Public school = 1 | 0.21 | −0.31 |
Had to repeat a year or more in school = 1 | −0.32*** | −0.2 |
Perception regarding performance (top student = 1) | 0.44 | 0.39** |
Perception regarding effort (large = 1) | 0.09 | 0.27** |
Constant | −0.17 | −0.01 |
Wald test of indep. eqns. | ||
Prob > χ 2 = 0.3111 | ||
Rights and permissions
About this article
Cite this article
Castro, J.F., Yamada, G. & Arias, O. Higher education decisions in Peru: on the role of financial constraints, skills, and family background. High Educ 72, 457–486 (2016). https://doi.org/10.1007/s10734-016-0040-x
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10734-016-0040-x