Subjective Earnings and Academic Expectations of Tertiary Education in Colombia

Ensayos sobre política económica Volumen 36, Núm. 86 • Edición Junio de 2018 159 Subjective Earnings and Academic Expectations of Tertiary Education in Colombia Luis Fernando Gamboa Paul Andrés Rodríguez-Lesmes* Article Info: Received 21 July 2017; accepted 22 January 2018 Abstract JEL Classification I25 J24 D84 Keywords: Subjective expectations Colombian education tertiary education We analysed the academic expectations and earning expectations of a nationally representative sample of Colombian students who were finishing their upper-secondary education (high school) and considering three potential scenarios: whether to finish their studies, enrol in a vocational career, or pursue a college degree. We found that these students’ earning expectations are correlated with local labour market wages. However, they expect earnings as university graduates that are significantly above the current observed earnings, which is not the case for the other two levels of education. We also found that earning and academic expectations (test scores and odds of attending college) are closely related to their family socioeconomic background, school and municipality characteristics, and even reflect aspects such as gender gaps or private school premiums. Finally, both academic and earnings expectations are related to actual realised test scores. Ingresos Subjetivos y Expectativas Académicas de la Educación Terciaria en Colombia Resumen Clasificación JEL I25 J24 D84 Palabras clave: Expectativas Subjetivas Educación en Colombia Educación terciaria * Analizamos las expectativas académicas y de ingresos de una muestra representativa de estudiantes colombianos en el último año de educación media secundaria, en el cuál consideramos tres posibles escenarios: terminar sus estudios, inscribirse en una carrera vocacional, u obtener un título profesional. Encontramos que los estudiantes tienen expectativas de ingresos correlacionadas con los salarios observados en el mercado laboral. No obstante, sus expectativas de ingreso en el caso de obtener un título profesional son notablemente superiores a los ingresos observados, lo que no ocurre en los otros dos posibles escenarios. También encontramos que las expectativas de ingreso y académicas (resultados de pruebas estandarizadas y la probabilidad de ir a una universidad) están relacionadas con las características socioeconómicas de sus familias, colegio y variables clave del municipio, reflejando aspectos como las brechas de género o el bono de estudiar en un colegio privado. Finalmente, los dos tipos de expectativas están relacionadas con los resultados obtenidos en las pruebas oficiales estandarizadas. https://doi.org/10.32468/espe.8601 Los autores pertenecen, en su orden, al Departamento de Economía de la Universidad Jorge Tadeo Lozano y a la Facultad de Economía de la Universidad del Rosario. Correos electrónicos: 160 Subjective Earnings and Academic Expectations of Tertiary Education in Colombia Luis Fernando Gamboa y Paul Andrés Rodríguez-Lesmes / Ensayos sobre Política Económica 86 (2018) 159–177 1. Introduction The purpose of this paper is to assess whether Colombian students in their last year of secondary school have academic and earnings expectations that relate to the labour market, and to the scores of standardised tests used for admission to colleges around the country. We use the case of Colombia, where higher education includes vocational (technical) and academic studies as alternative options. The understanding and practical application of basic principles of science and mathematics is carried out in the former, and its objective is the preparation of graduates for occupations that are classified above skilled crafts, but below scientific or engineering professions. Professional education places a major emphasis on the theories, understanding, and design of new ways to solve problems. People invest in education under uncertain schemes seeking non-pecuniary goals (Oreopoulos and Salvanes, 2009), but also improving their quality of life by maximising their earning potential. These concepts are summarised in the well-known human capital formation model (Becker, 1964; Mincer, 1974), where earnings expectations play a central role in investment choices. An important element to consider is the distinction between ex ante and ex post income variability. In some middle-income and developing countries, that individuals’ ex ante forecasts about their future income might be below the ex post income that they would obtain if they were to invest in education: Jensen (2010) found that teenagers in rural parts of the Dominican Republic do not expect important benefits from secondary education. In a different direction, Hastings et al. (2016) found that students who overestimate or underestimate the costs of college in Chile tend to make “mistakes” regarding human capital formation. Jensen’s results differ from the findings from developed countries. While Dominitz and Manski (1996) found that students know about the existence of a college earnings premium, they demonstrate substantial heterogeneity and commonly overestimate (ex post) returns. However, in the case of Duke University students, they are aware of the relative pecuniary benefits of their majors (Arcidiacono et al., 2010). There is also evidence about the role played by externally implemented tests on expected returns, choices, and student motivations (Stinebrickner and Stinebrickner, 2012; Wiswall and Zafar, 2015; Zafar, 2011; Zafar, 2013; Reuben et al., 2017; Sequeira et al., 2016). Based on this evidence, several information-based policy experiments have been conducted around the world. Information on labour market wages around the world has been used in China (Loyalka et al., 2013), Colombia (Bonilla et al., 2015), Finland (Pekkala et al., 2015), India (Sequeira et al., 2016), Mexico (Avitabile et al., 2015) and the United States of America (Fryer, 2013). Others have focused instead on knowledge about costs and financial constraints; for example, Hoxby and Turner (2013) in the United States of America, and McGuigan et al. (2015) in England. Another methodology, based on reference points, has been used in Madagascar (Nguyen, 2013). The results from these interventions are mixed, with some students updating their beliefs, but often without significant changes in their actual choices. However, there is evidence that another type of intervention, such as conditional cash transfers, might improve students’ aspirations (García et al., 2016). It is important to mention that future earnings expectations may differ from contemporaneous data because of several reasons. This is due to either having private information to explain why their idiosyncratic beliefs differ from the observed mean, or to having incorrect information about such a mean. Systematic differences arise if students consider that economic conditions will be different in the future if there is a generalised error i (...truncated)