second workshop on education economics

Second Workshop on Education Economics Maastricht, March 23-24, 2016 Top Institute for Evidence Based Education Research

Book of Abstracts

This workshop on Education Economics seeks to explore the use of economic and econometric techniques to study educational issues.

The workshop is organized by the research centre Top Institute for Evidence Based Education Research (TIER) at Maastricht University and the Leuven Economics of Education Research (LEER) at KU Leuven.

Book of Abstracts – Workshop on Education Economics

MAASTRICHT, MARCH 23-24, TOP INSTITUTE FOR EVIDENCE BASED EDUCATION RE SEARCH ─ 1



O V E R V I E W P A R A L L E L S E S S I O N S

Room 0.009 – Education Labour Market Room 1.008 – Confidence and behavior Room 2.006 – Instruction Time

Wednesday, March 23 10.25-12.25

1 The Labour Market Position of Early School-leaving Youth: are the prospects really bad? Sofie Cabus

The Effect of Multigrade Classrooms on Student Behavior. Reza Sattari

New Evidence on the Effect of Computerized Individualized Practice and Instruction on Language Skills. Joris Ghysels & Carla Haelermans

2 Biligual Schooling and Earnings: Evidence from a Language-in-Education Reform. Lorenzo Cappellari & Antonio Di Paolo

Motivation to Learn: Confidence in Levels versus Confidence in Gains. Mira Fischer and Dirk Sliwka

High School duration and its effects on university education. Evidence from the G8 reform. Jan Marcus & Vaishali Zambre

3 Mismatch between Education and the Labour Market in the Netherlands: Is it a Reality or a Myth? – The Employers Perspective. Sofie Cabus & Melline Somers

Children Left Behind: Self-confidence of Pupils in Competitive Early Tracking Environments. Miroslava Federičová, Filip Pertold, & Michael Smith

The Heterogeneous Effects of a Compulsory school leaving Age Raise. Anna Adamecz

4 Does it Pay to Move? Returns to Regional Mobility to the First Job for German Graduates. Michael F. Maier & Maresa Sprietsma

The interaction between dropout, graduation rates and quality ratings in universities. Eline Sneyers & Kristof De Witte

The Attractiveness of Programs in Higher Education: An Empirical Approach. Ferdi Widiputera, Kristof De Witte, Wim Groot & Henriëtte Maassen van den Brink

Room 0.009 – Efficiency in Education Room 1.008 – Education-labour market Room 1.028 – Student achievement

Wednesday, March 23 14.40-16.10

1 Efficiency and competition in secondary schools in the UK. Laura Lopez Torres, Jill Johnes & Caroline Elliott

Youth unemployment: the role of cognitive skills and labor market institutions. Margarida Rodrigues

Thinking Smart or Thinking Fast? Experimental Evidence on Framing Effects in Elementary Schools. Valentin Wagner.

2 Productivity Growth of Norwegian Institutions of Higher Education 2004 – 2013. Dag Fjeld Edvardsen. Finn R. Førsund & Sverre A. C. Kittelsen

The incidence of and returns to overeducation: PIAAC evidence on the G7. Geraint Johnes

The Influence of Class size on Academic Underachievement. An advanced Stochastic Frontier Approach. Deni Mazrekaj, Kristof De Witte & Thomas Triebs



3 A Unified Productivity-Performance Approach, with an Application to Secondary Schools in the Netherlands. Laurens Cherchye, Kristof De Witte & Sergio Perelman

The effects of the 1999 Polish educational reform on labour market outcomes. Drucker, Luca Flóra & Dániel Horn

Secondary school track choice and local characteristics. Kłobuszewska Małgorzata & Rokicka Magdalena

Room 0.009 Grading Room 1.008 – Education-labour market

Wednesday, March 23 16.40-17u40

1 Does Truancy Cause Bad PISA Results? What can be Learned on the Effect in Presence of Measurement Error. Stefanie Hof & Christine Sälzer

How does maternal labor supply respond to changes in children’s school schedule? Emma Duchini & Clémentine Van Effenterre

2 The Impact of Early Grading on Academic Choices: Mechanisms & Social Implications. Luca Facchinello

The effects of higher teacher pay on teacher retention. Sander Gerritsen, Sonny Kuijpers & Marc van der Steeg

Room 0.009 School management Room 1.008 - Inequality Room 1.028 - Gender

Thursday, March 24 9.55-12.55

1 School district consolidation: Scale economies and the optimal scale of education. Fritz Schiltz & Kristof De Witte

Where are the resilient schools and what makes the school resilient. Jean Hindriks & Mattéo Godin

The Effect of Adaptive versus Static Practicing on Student Learning - Evidence from a Randomized Field Experiment. Chris Van Klaveren, Ilja Cornelisz & Sebastiaan Vonk

2 Italian school principals’ managerial behaviors and students’ test scores: an empirical analysis. Tommaso Agasisti, Patrizia Falzetti & Mara Soncin

Short- and long-term effects of elite high schools. Ohto Kanninen & Mika Kortelainen

Fertility Timing and Education. Jonathan James & Sunčica Vujić

3

Private schools and student achievement. Ebrahim Azimi, Jane Friesen & Simon Woodcock.

Estimating the impact of health on NEET status. Daniel Gladwell, Gurleen Popli, & Aki Tsuchiya

Education and skill matching in the European Labour Markets. Some preliminary evaluations from a multilevel analysis. Laura Chies & Grazia Graziosi



4 Parents, Schools and Human Capital Differences across Countries. Marta De Philippis & Federico Rossi

The effect of parental involvement on digital practice for math- and language of secondary students. Carla Haelermans & Joris Ghysels

Do Rewards Reduce Effort? Evidence From University Exams. David Kiss, Malte Sandner & Daniel Schnitzlein

5 The demand for autonomous schools. Marco Bertoni, Stephen Gibbons & Olmo Silva

Enrollment and degree completion in higher education without admission standards. Koen Declercq & Frank Verboven



2016 Wednesday, March 23

8.45-9.15 Registration and coffee | Court Yard 9.15-9.30 Welcome | Room 0.009

Kristof De Witte (TIER - Maastricht University, LEER – KU Leuven)

Wim Groot (TIER – Maastricht University) 9.30-10.15 Keynote speaker:

Sandra McNally (London School of Economics) | Room 0.009

Title: Teaching and Learning Literacy

with Stephen Machin and Martina Viarengo

Abstract: A significant number of people have very low levels of literacy in many OECD countries. Many economic studies show high labour market returns to literacy – although the benefits extend well beyond the labour market. However, it is not clear what teaching strategies are most useful for actually improving literacy. It is hugely controversial amongst educationalists and little studied by economists. Research evidence in Scotland prompted a national change in the guidance given to schools in England in the mid-2000s about how children are taught to read. We conceptualise this as a shock to the education production function. We use the phasing in of intensive support to schools to estimate the effect of the new teaching strategy on children’s educational attainment. Using longitudinal data, we can track effects throughout primary school. We also estimate effects for those with a higher propensity to struggle with reading at an early age (i.e. those from poor backgrounds and non-native speakers of English). We find there to be short-term effects of the teaching technology (‘synthetic phonics’) at age 5 and 7 that appear to fade out by the age of 11, except for those groups with a higher propensity to struggle with reading – where we find evidence of a persistent effect.

10.15-10.25 Room change 10.25-12.25 Parallel sessions – 1



Session 1-1 | Room 0.009

• The Labour Market Position of Eerly School-leaving Youth: are the prospects really that bad? Sofie Cabus (TIER Maastricht University)

Since the Lisbon agenda targets set in the year 2000, early school-leaving is high priority on many policy agendas in the OECD. While previous studies mainly focused on the determinants of school dropout or dropout prevention measures, this research collects evidence on the labour market position of early school-leaving youth in the Netherlands. It is researched in particular whether school dropouts increasingly experience job mismatch on the labour market compared to well-educated peers in terms of: the duration of unemployment spells (e.g. first job, spells between jobs); wage differentials; and ultimately job content. The research also distinguishes the short-term from the long-term labour market consequences. The study uses a large and complete administrative dataset of Statistics Netherlands (CBS). The data consist of detailed information on the labour market position of school dropouts and graduates over the past 10 years. The research heavily relies on cohort study life-cycle analyses. Within cohorts, comparison will be made between the labour market outcomes of school dropouts and graduated peers. While the decision to quit school early is endogenous with labour market outcomes, we estimate the effects of school dropout on: (1) unemployment spells by using a timing of the events approach; (2) wage differentials by using Heckman selection and Rubin matching models; and (3) job contents by using multinomial logit models. We will also search for policy changes aiming at school-leaving youth in order to find proper instruments for school dropout, but up till now we did not find those yet. Based on previous research on vocational school-leaving youth (Cabus and De Witte, 2011; Cabus and Haelermans, 2015), we expect to find substantial differences between short-term and long-term labour market outcomes for different educational levels. In the short-term, financial incentives and other opportunity costs will attract schoolchildren in labour market positions and into early school-leaving. However, in the long-term well-educated peers are expected to be better-off in terms of competitiveness on the labour market and financial rewards. Furthermore, previous research is not clear on the differences in job contents and corresponding incidence of mismatch between education and the labour market for school dropouts and graduates.

• Bilingual Schooling and Earnings: Evidence from a Language-in-Education Reform* Antonio Di Paolo (Universitat de Barcelona) With Lorenzo Cappellari (Università Cattolica Milano)

We exploit the 1983 language-in-education reform that introduced Catalan alongside Spanish as medium of instruction in Catalan schools to estimate the labour market value of bilingual education. Identification is achieved in a difference-in-differences framework exploiting variation in exposure to the reform across years of schooling and years of birth. We find positive wage returns to bilingual education and no effects on employment, hours of work or occupation. Results are robust to education-cohort specific trends or selection into schooling and are mainly stemming from exposure at compulsory education. We show that the effect worked through increased Catalan proficiency for Spanish speakers and that there were also positive effects for Catalan speakers from families with low education. These findings are consistent with human capital effects rather than with more efficient job search or reduced discrimination. Exploiting the heterogeneous effects of the reform as an instrument for proficiency we find sizeable earnings effects of skills in Catalan.



• Mismatch between Education and the Labour Market in the Netherlands: Is it a Reality or

a Myth? – The Employers’ Perspective Melline Somers (TIER Maastricht University) With Sofie Cabus (TIER Maastricht University)

Almost 50 percent of Dutch employers reported mismatch between skill demand and skill supply in 1991. In 2011, this rate was only 25 percent. One potential explanation for this decline is that employers benefitted from formal schooling trends that developed in response to the increasing demand for higher-educated graduates. This hypothesis is tested in a two step empirical framework using data from the Dutch Labour Demand Panel that covers the period between 1991 and 2011 and consists of 7,451 unique companies. First, we show that an increase in formal schooling of the labour force with one month increases the schooling level of the staff with almost 0.3 months. Second, we link these changes in the schooling level of the staff, owing to a higher-educated labour force, with mismatch. The results indicate that a one-month increase in the companies’ workforce average schooling level decreases the probability that companies report mismatch with -2.7 percentage points.

• Does it Pay to Move? Returns to Regional Mobility to the First Job for German Graduates Maresa Sprietsma (Centre for European Economic Research, ZEW Mannheim). With Michael F. Maier (Centre for European Economic Research, ZEW Mannheim).

We estimate the contemporary wage returns to regional mobility after graduation from university. The mobility decision at the start of the career is of high relevance for future wage and employment prospects. In effect, the majority of German graduates are mobile only at this point of their career. To control for self-selection, we jointly estimate the decision to move after secondary school and after graduation from university applying a bivariate probit estimation. Furthermore, we use the availability of places to study in the region of secondary school as a source of exogenous variation for the decision to be mobile. We find positive significant returns to early regional mobility for German graduates.


• The Effect of Multigrade Classrooms on Student Behavior Reza Sattari (Department of Economics Simon Fraser University) Amid growing evidence of the importance of non-cognitive skills for both cognitive skill development and long-term outcomes, understanding the effect of education policies on non-cognitive skill formation is of increasing interest. This paper provides the first quasi-experimental evidence of the effect of multi grade classes on non-cognitive skills. I exploit strictly enforced class size caps accompanied by centralized funding rules to generate IV estimates of this effect, using custom survey data administered to over 15000 parents of Kindergarten and Grade 1 students, linked to publicly available administrative data on multi grade classes. I find that placing children in multi grade classes causes a significant increase in peer relationship problems and hyperactivity relative to placement in single grade classrooms.

• Motivation to Learn: Confidence in Levels versus Confidence in Gains

Mira Fischer (University of Cologne) With Dirk Sliwka (University of Cologne)



Motivational beliefs are held to be an important determinant of success in education. Whereas in educational settings it is sometimes claimed that optimistic beliefs about own skills affects learning positively, economists have mostly pointed at the detrimental effects of high confidence in own skills. There also seems to be disagreement about whether praise for performance, effort or progress is best to raise confidence and motivation to learn. A straightforward conjecture is that much of the disagreement in the popular discourse about the relationship between feedback, confidence, and performance is caused by the tendency to lump together different types of beliefs into the category of “confidence”. Different types of feedback may influence different confidence beliefs, some of which raise motivation to exert effort whereas others do not. Indeed, the literature in psychology indicates that there is mixed evidence on the effects of different types of feedback on performance. Some types of praise are more likely to raise the perceived level of ability whereas others are more likely to raise the perceived effectiveness of learning effort and we should not expect that both changes in beliefs will have the same effect on motivation to learn. Economists who have studied confidence in experimental settings have mostly conceived of it to refer to a belief about one's ability relative to one's true ability or relative to others' ability and have focussed on the effects of “overconfidence”, or believing that one's ability is greater than it actually is. This study, however, focusses on the effects of raising or lowering confidence on motivation. The purpose of this paper is to analyze the separate role of confidence about prior knowledge and confidence in the ability to learn for the incentives to make human capital investments.

• Children Left Behind: Self-confidence of Pupils in Competitive Early Tracking

Environments Filip Pertold (CERGE- EI) With Miroslava Federičová and Michael Smith Early-tracking systems, naturally divides many 5th grade classes into two groups: pupils preparing for exams to enter better schools and everyone else, who do not compete for selective schools. We show that this environment has a detrimental effect on the self-confidence of pupils in mathematics who do not apply for selective schools but have peers in their classroom who do apply. Girls who do not apply experience drop in confidence by 11 percent if they have four applicants among classmates and this effect is even larger if the applicants are successful in admission process. We focus on self-confidence in mathematics as an outcome variable because the literature suggests it is directly linked to pupils’ motivation to study STEM (science, technology, engineering and mathematics) fields as well as subsequent educational achievement. Our findings are also relevant to the child psychology literature, which has shown that children around the age of 11 have a high degree of inequality aversion and the presence of more successful peers can induce an increase in negative emotions, such as envy.

• The interaction between dropout, graduation rates and quality ratings in universities. Eline Sneyers (TIER Maastricht University) Kristof de Witte (TIER Maastricht University – KU Leuven) This paper investigates in a non-parametric framework whether academic programs maximize their student graduation rates and program quality ratings given the first year student dropout rates. In addition, it explores what institutional and program characteristics explain this interaction. The results show a large variation in how academic programs are able to deal with the selective nature of first-year dropout. Nevertheless, we can accurately explain the variation among programs by program and institutional characteristics. It seems that universities can maximize the relation between first-year dropout, graduation rates and



quality ratings in several ways: (1) by improving student program satisfaction, (2) by better preparing certain groups of students for higher education, (3) by supporting male students, (4) by supporting ethnic minority students, (5) by attracting older staff, and (6) by strengthening the selective nature of the first-year (i.e., increasing the academic dismissal policy threshold).

Session 1-3 | Room 2.006 – Instruction time

• New Evidence on the Effect of Computerized Individualized Practice and Instruction on Language Skills Joris Ghysels (TIER Maastricht University) With Carla Haelermans (TIER Maastricht University) This paper provides new evidence on the effect of computerized individualized practice and instruction on spelling. An individually randomized experiment among 7th grade students in the Netherlands is developed to study the effect of an adaptive IT homework tool (intelligent tutoring system, ITS) on spelling performance. Using an IV-approach to control for actual use of the IT-tool, we show that there are small positive effects of practicing with an adaptive IT-tool for spelling for 7th grade students. Effects are largest for low-performing students. The additional homework time investment by students proves to be cost effective as well, implying efficient use of the tool by the students.

• High school duration and its effects on university education.

Evidence from the G8 reform Vaishali Zambre (DIW Berlin) With Jan Marcus (DIW Berlin) A major education reform in Germany reduced the number of years required for high school graduation by one year, while keeping total instruction time constant. The main goal of this so-called G8 reform was to decrease the age at labor market entry without decreasing the human capital of graduates, thereby addressing challenges of demographic change. While not intended by policy makers, the higher learning intensity experienced during high school may also affect other education outcomes, even after high school. For a better understanding of the overall consequences of the controversial G8 reform, we investigate whether the reform affected students’ choice to enroll in college as well as the timing of their enrollment. The fact that the federal states introduced the reform at different points in time produces a natural experiment. We exploit this variation in time and across states using a difference-in-differences approach. We use administrative data (”Studierendenstatistik”) that covers all students enrolled in any German university between 2002 and 2013 (over 1.57 million students). Our results show that the G8 reform did not only decrease college enrollment by about 4 percentage points, but also resulted in a delayed enrollment. This result is robust in a broad range of sensitivity tests and persists over time. The results of our study are not only informative for the German context but also for policy makers in many other OECD countries, trying to increase the number of active labor market participants in order to address the challenges of an aging society.



• The Heterogeneous Effects of a Compulsory school leaving Age Raise. Anna Adamecz-Völgyi (Central European University) This paper examines the effect of a rise in compulsory school leaving (CSL) age from 16 to 18 in Hungary on the children of uneducated parents. The new regulation got into force in the case of those who started primary school in September 1998, thus, theoretically, it started to be binding for those born in June 1991. This policy change offers a natural cutoff between those born in May vs. June 1991 for a regression discontinuity design-based identification strategy. However, in practice, a reasonable share of parents is non-compliant as they tend to have a preference for holding their children back for a year from starting primary school, and this effect is the strongest right before the cutoff point. This paper identifies the causal impact of the policy change in such a fuzzy regression discontinuity design where school starting choice is instrumented by a date-of-birth-determined treatment status in a twosample two stage least squares (TS2SLS) setup. First results show that longer compulsory schooling increases the probability of earning a higher-tier secondary degree which brings substantial gains on the labor market. The impact is realized on the intensive margin by inducing students to earn a higher-tier secondary degree instead of a lower-tier one. The legislation change had no effect on the extensive margin: it did not increase the probability of earning a secondary degree in general. Estimating 2SLS instead of intention-to-treat effects unfolds that estimations not taking school entering non-compliance into account may underestimate the effect of such policy changes.

• The Attractiveness of Programs in Higher Education: An Empirical Approach. Ferdi Widiputera (TIER Maastricht University) With Wim Groot, Henriëtte Maassen van den Brink and Kristof De Witte

We investigate the relationship between the attractiveness of a study program in higher education and the location of the institution offering the program (including the distance), program (i.e., type of program, type of education), city characteristics (i.e., number of city residents, room rental cost, etc) and other characteristics (i.e., competition). This study focusses on the program level and uses panel data from 2008 to 2013. The analyses are based on a data collected by the Dutch Ministry of Education, comprising more than 1300 study programs offered by 50 institutions. The results show significant differences by distance and other characteristics. The results suggest that distance and competition significantly does not influence the attractiveness of a program. The distance to the nearest institutions which offers the same program and the competition between similar programs in cities does not affect a student’s decision for a program. Our findings also indicate that other city characteristics (i.e., the students’ available room, the number of pubs, the number of museums and the students’ culture affect the attractiveness of a program.

12.25-13.45 Lunch | Court Yard 13.45-14.30 Keynote speaker:

Hessel Oosterbeek (University of Amsterdam) | Room 0.009

Title: The performance of school assignment mechanisms in practice



with Monique de Haan, Pieter Gautier & Bas van der Klaauw Abstract: Theory points to a potential trade-off between two main school assignment mechanisms: Boston and Deferred Acceptance (DA). While DA is strategy-proof and gives a stable matching, Boston might outperform DA in terms of ex-ante efficiency. We quantify the (dis)advantages of the mechanisms by using information about actual choices under (adaptive) Boston complemented with survey data eliciting students' school preferences. We find that under Boston around 8% of the students apply in the first round to another school than their most-preferred school. We compare allocations resulting from Boston with DA with single tie-breaking (one central lottery; DA-STB) and multiple tie-breaking (separate lottery per school; DA-MTB). DA-STB places more students in their top- n schools, for any n, than Boston and results in higher average welfare. We find a trade-off between DA-STB and DA-MTB. DA-STB places more students in their single most-preferred school than DA-MTB, but fewer in their top- n, for n ≥ 2. Finally, students from disadvantaged backgrounds benefit most from a switch from Boston to one of the DA mechanisms.

14.30-14.40 Room change 14.10-16.10 Parallel sessions – 2 Session 2-1 | Room 0.009

• Efficiency and competition in secondary schools in the UK Jill Johnes (University of Huddersfield) With Laura Lopez Torres & Caroline Elliott Government policy in the UK has over time encouraged schools in the public sector to become more competitive in order to reap savings from increased efficiency. In order to test whether it is indeed the case that competition leads to high technical efficiency we examine operational efficiency in the independent schools sector in the UK which has always operated in a competitive environment. Findings regarding the efficiency of schools in this sector are of interest to both independent and state schools as the latter move into a more competitive environment. Efficiency is measured using a production function approach that considers schools to be ‘firms’ which convert ‘inputs’ into ‘outputs’. We use a bootstrapped data envelopment analysis (DEA) approach to estimate the production function and derive estimates of efficiency. This first stage in which the production function and efficiency of the sector are estimated is just the starting point of our work. In the second stage, we investigate the specific determinants of each school’s performance, including levels of competition amongst schools. We use a panel data random effects estimation approach with clustered standard errors to assess the significance of these variables, with specific focus on the competition variables.. The effects of competition on efficiency are particularly interesting (and perhaps contrary to expectation) and lead to policy recommendations which will be of interest to those in government with a responsibility for education and also to managers and researchers in education sectors.



• Productivity Growth of Norwegian Institutions of Higher Education 2004 – 2013 Finn R. Førsund (Department of Economics, University of Oslo) With Sverre A.C. Kittelsen and Dag Fjeld Edvardsen Studies of productivity of institutions of higher institutions is of interest for two main reasons; education is an important factor of productivity growth for the economy, and in countries where higher education is funded by the public sector the effectiveness of spending the resources, i.e. accountability, is of key interest. Six variables are specified and measured yearly; two inputs, faculty and administrative staff, and four outputs, study points obtained for lower and higher education, publishing points for peer refereed research publications, and Ph.Ds granted. A bootstrapped Malmquist productivity index is used to calculate productivity development for Norwegian institutions of higher education over the 10 year period 2004-2013. The confidence intervals got from the bootstrapping allows part of the uncertainty of point estimates to be revealed. Aggregate development of productivity is shown but also the productivity development of the 44 individual institutions constituting a panel. Figures are used to illustrate the heterogeneity of the development of productivity for each unit including the confidence intervals for three cross sections and one from the first to the last year. The main result is that the majority of institutions have had a positive productivity growth over the total period. Comparing change in total labour, size of the units and productivity development the smallest units show the strongest heterogeneity, but there is also a considerable difference in change in the use of labour of large units compared with the spread in productivity.

• A Unified Productivity-Performance Approach, with an Application to Secondary Schools in the Netherlands. Sergio Perelman (University of Liège) With Laurens Cherchye (KU Leuven) & Kristof de Witte We introduce a novel and simple diagnostic tool to improve the performance of public services (e.g. in health, education, utilities and transportation). We propose a method to compute performance/productivity ratios, which can be applied as soon as data on production units’ outcomes and resources are available. These ratios have an intuitive interpretation: values below unity indicate that better outcomes can be attained through weaker resource constraints (pointing at scarcity of resources) and, conversely, values above unity indicate that better outcomes can be achieved with the given resources (pointing at unexploited production capacity). We demonstrate the practical usefulness of our methodology through an application to secondary schools in the Netherlands. In this application, we also account for outlier behavior and environmental effects by using a robust nonparametric estimation method. Our empirical results indicate that in most cases schools’ performance improvement is a matter of unexploited production capacity, while scarcity of resources is a lesser issue.

Session 2-2 | 1.008

• Youth unemployment: the role of cognitive skills and labour market institutions. Margarida Rodrigues (Institute for Employment Research (IAB)

. The aim of this paper is to assess to what extent countries’ youth unemployment levels are associated with youth human capital, and how this association interacts with different labor market institutions. Cognitive skills scores from PISA international test assessments are used as a proxy for human capital. These are combined with youth unemployment ratio of 23 European countries for the years 2001-2011. Using a fixed effects panel data approach



we find that the higher the cohorts’ cognitive skills at the age of 15, the lower is that cohorts’ aggregate unemployment later on. As expected this association is weaker if the minimum wage has higher coverage and is more centralized and if employment protection legislation is stricter.

• The incidence of and returns to overeducation: PIAAC evidence on the G7 Geraint Johnes (Lancaster University)

PIAAC data are used to evaluate the extent of overeducation in G7 countries. Incidence of overeducation is seen to vary systematically with a number of demographic characteristics. The impact of overeducation on remuneration is then estimated using quantile regression. This impact is observed to be minor, suggesting that while some individuals have qualifications in excess of those required to undertake their job, their additional human capital is nonetheless rewarded. Care therefore needs to be taken in interpreting measures of overeducation.

• The effects of the 1999 Polish educational reform on labour market outcomes.

Daniel Horn (CERSHAS) With Drucker & Luca Flora

The 1999 Polish educational reform is often considered successful as the results of the Polish students on the OECD PISA tests have improved significantly since the introduction of the new system. Due to this reform the previous 8-year undivided comprehensive education was extended to 9 years, core curricula were introduced and the examination, admission and assessment systems were changed. It has been argued before that this longer comprehensive education helped the test performance of worse performing students; hence increasing average performance and decreasing inter-school variation of test scores. However the lack of reliable impact assessment on the long-run effects of this reform is alarming. In this paper we aim to fill this gap by looking at the causal effects of the reform. By comparing the labour market outcomes of the pre- and post-reform cohorts, we find a non-negligible and positive effect. We look at outcomes such as participation in education, probability of employment and wages. Using data from the EU-Statistics on Income and Living conditions, pooling the waves between 2005 and 2013 and taking only the 20-27 year-olds, we generate a quasi-panel of observations. In all estimations we control for year of survey, age, experience, and labour market entry characteristics: a quasi regression discontinuity method. We find evidence that the reform was successful on the long-run. Although there is no difference in the participation rates in the different levels of education between the two groups, the post-reform group is more likely to be employed and that they also earn higher wages. On average the treatment group is more than 2% more likely to be employed, which effect is likely to be driven by the lowest educated (cca 6%). The post-reform cohort is also more likely to earn more: we find a 3-4% difference in real wages, which is also more pronounced for the lowest educated.


• Thinking Smart or Thinking Fast? Experimental Evidence on Framing Effects in Elementary Schools Valentin Wagner (University of Dusseldorf))

The aim of this paper is to test the motivational power of different grading schemes -different framing of allocation of points - on pupils' performance in a mathematical multiple-choice test. We present results of a field experiment on 1377 pupils from 20 elementary



schools in Germany which were randomized into one control group and two treatment groups. Pupils in the Control Group started the test with 0 points but could gain 4 points for a correct answer and 2 points if the answer was skipped. No points were deducted if the answer was wrong. In the Loss Treatment, pupils were given the maximum number of points (40 points), but 4 points were deducted for a wrong answer and pupils lost 2 points if a question was skipped. In the Negative Treatment, pupils started with -20 points and could gain a maximum of +20 points. Again, 4 points were allocated for a correct answer, 2 points if the answer was skipped and 0 points for a wrong answer. We find significant and positive effects on the number of correct answers in both treatments for boys as well as girls. Furthermore, we can identify two different mecanisms that drive our treatment effects. In the Loss Treatment, pupils seem to think faster than pupils in the control group as they answer significantly more questions but the auccuracy -share of correct answers on all given answers - does not change. In contrast, pupils in the Negative Treatment seem to think smarter than pupils in the control group as the number of answered questions does not change but the accuracy significantly increases. As we received pupils' past midterm grade in math, we can additionally analyze the effectiviness of framing effects separately for low- and high-ability pupils.

• The Influence of Class size on Academic Underachievement. An advanced Stochastic Frontier Approach Deni Mazrekaj (KU Leuven) With Kristof De Witte (KU Leuven, Maastricht University) and Thomas Triebs (CESifo) The literature has traditionally measured underachievement as the difference between capacity test scores, such as IQ, on the one hand, and achievement test scores or school grades on the other. This paper questions the current empirical practice and suggests an advanced Stochastic Frontier approach to separate pupil effects, persistent underachievement and time-varying underachievement. Drawing on Flemish data from kindergarten until grade three we show that overall underachievement amounts to 46% and is mainly driven by permanent factors, making it likely to persist over time. Moreover, we present evidence that an increase in class size by ten per cent may lead to a corresponding increase in underachievement by 0.79 percentage points. This effect appears to be non-monotonic and is likely to occur solely above the class size of eight students. Below this threshold, larger classes may actually reduce underachievement. The results further indicate that boys underachieve twice as much as girls and that larger classes decrease underachievement for gifted students, while increasing underachievement for the non-gifted students. We argue that a class size of eight students yields highest benefits, reducing underachievement by 36%. However, the most cost-effective policy is a class size reduction to ten students, yielding a net benefit of around 2 891 euros per pupil.

• Secondary school track choice and local characteristics. Magdalena Rokicka (IBE Educational Research Institute) With Malgorzata Klobuszewska This paper addresses the issue of a secondary school track choice in the perspective of the local educational system in Poland. We assume that an upper secondary school choice is influenced both by characteristics of a child and its household and by characteristics of a county they live in. Although there is a common core curriculum for each type of school in Poland, the local authorities are in charge of governance, maintaining the upper secondary schools located within their borders. Also, differences in the labour market affect the potential return on educational path. This creates disparity between different regions mainly



in regards to the school provision, accessibility, and their coverage. This paper examines to which extent the school track choice between vocational and general education can be attributed to variation of personal characteristics and to which extent to variation of the local conditions. The school choice at this stage is of a crucial importance as it determines the further educational path, and future labour market prospects. To address this issue we use data collected within a project BECKER during 2012/2013 in several Polish counties. We applied a multilevel method to address the impact of household characteristics on the individual level as well as the local characteristics on the county level. As we made distinction between three different types of secondary schools: basic vocational, secondary vocational, and general secondary we model the choice using a functional form of multinomial logistics regression.

Our results indicate that both child’s characteristics such as sex and school performance and characteristics of household (parents’ education, material situation) influence a school track choice. We also found that there is statistically significant variation in school track choice on the county level. Local unemployment rate is statistically significant determinant of school track decision and it also explains a great variation between counties. This finding is important in the light of EU regional cohesion policy aiming at improving economic situation of regions and diminishing regional disparities also in regards to education.

16.10 – 16.40 Break and refreshments | Courtyard

16.40 – 17.40 Parallel sessions – 3


• Does Truancy Cause Bad PISA Results? What can be Learned on the Effect in Presence of Measurement Error. Stefanie Hof (Swiss Coordination Center for Research in Education) With Christine Sälzer (ZIB Center for International Student Assessment) Truancy is often thought of as a problematic student behavior related to low academic achievement. Research is contradictory and sparse. From a methodological point of view, the fact that the group of truant students is prone to be a self-selective one needs to be taken into account when studying truancy. Further, there is reason to assume that truancy is sometimes misreported in terms of exaggerated frequencies or denied behavior, meaning that students’ responses to questionnaire items may involve both false positive and false negative classifications of students as truants or non-truants (e.g. Siziya, Muula, & Rudatsikira, 2007). Therefore, we explored the causal impact of truancy on students’ academic achievement in mathematics while accounting for both non-random selection into treatment and measurement error. We address these identification problems with a partial identification framework. Instead of obtaining point estimates, this nonparametric method of analysis yields bounds around the average treatment effect (ATE) and is thus more credible. Moreover, the analysis drops the probably rather unrealistic assumption of a linear and homogenous effect of intentionally skipping mathematics classes on students’ performance in the PISA mathematics test. We impose several weak assumptions concerning the nature of selection and the measurement error. In our analyses, we use more than 9,000 students from the grade-based PISA 2012 national sample in Germany. The identified bounds are still quite large and include zero. However, the impact of misreporting is profound. If as little as two percent of the sample misreport their truancy behavior, the direction of the ATE even under exogenous selection is not clearly negative



anymore. Thus, it is obvious that the association between truant behavior and student academic achievement is not robust to even small amounts of misreporting.

• The Impact of Early Grading on Academic Choices: Mechanisms and Social

Implications. Luca Facchinello (Stockholm School of Economics) Does early grading affect educational choices? To answer this question, I exploit a curriculum reform which postponed grade assignment in Swedish compulsory schools. The staggered implementation of the reform allows me to identify short- and long-term effects of early grading, for students with different academic ability and socioeconomic status (SES). When graded early on, high-ability students (especially if high-SES) exhibit higher grades in compulsory school, and are more likely to choose academic courses. Low-ability students react in the opposite way, with particularly negative reactions among low-SES students. High school attainment increases for high-ability low-SES students; college attainment decreases for low-ability low-SES students. None of these effects carry over to the labor market. This suggests that early grades improve the match between early education choices and academic ability, and reduce over-investment in education. I show that the short-term effects are consistent with predictions from a learning model in which children are uncertain about academic ability, have different priors depending on SES, and use grading information to re-optimize educational choices. I find no evidence of demotivating effects for low-ability students, an alternative mechanism through which grades might affect education choices, and the main motivation behind the grading reform.


• How does maternal labor supply respond to changes in children’s school schedule? Emma Duchini (Universitat Pompeu Fabra) With Clémentine Van Effenterre (Paris School of Economics) In this paper we analyze mothers’ labor supply response to a reorganization of children’s

school schedule. Until 2013, French children between 2 and 11 years old had their class time spread over 4 days and they did not go to school on Wednesday. In 2013 a national reform shortened each school day by an average of 45 minutes and reallocated the resulting three hours to Wednesday morning. We look at the impact on mothers’ labor supply exploiting variation in the implementation of this reform over time and across the age of the youngest child. We provide evidence of a reallocation of working hours over the week and no effect on the total number of hours worked. Overall, these results suggest that even in a context of high female labor force participation the organization of children’s time still affects mothers’ employment decisions.

• The effect of higher teacher pay on teacher retention.

Marc van der Steeg (Ministry of Education) With Sonny Kuijpers and Sander Gerritsen This paper investigates the effects of higher teacher pay for secondary school teachers on their teacher retention decision and enrollment in additional schooling. We exploit variation in teacher pay induced by the introduction of a new remuneration policy. This policy provided schools in an urbanized region with extra funds to place a larger share of teachers in a higher salary scale. We exploit this policy in an IV-setup to estimate the effects of higher teacher pay on our outcomes. The main finding is that we find no effects of higher teacher pay on the probability to stay in the teaching profession. The policy



however succeeded in keeping a slightly larger share of teachers in the targeted region. In addition, our findings suggest that the policy increased teachers’ enrollment in bachelor or master degree programs from 2.3% to 3.2%. This finding is consistent with the setup of the policy in which one of the criteria for placement in a higher salary scale is that teachers would obtain extra qualifications or gain extra expertise.

17.40 End of day 1 19.30 Dinner at the restaurant: ‘Au Coin Des Bons Enfants,

Ezelmarkt 4, Maastricht



2016 Thursday, March 24

8.45-9.00 Registration | Court Yard 9.30-10.00 Keynote speaker:

Daniele Checchi (University of Milan) | Room 0.009

Title keynote: Publish or Perish? Incentives and Careers in Italian Academia

Abstract: We derive a model of effort in the presence of career concern based on the multi-unit all-pay auction, and closely inspired by the Italian academic environment. In our model, the number of posts, the number of applicants, and the relative importance of the criteria for promotion determine academics’ effort and output. Because of the specific characteristics of Italian Universities, where incentives operate only through promotion, and where all appointment panels are drawn from strictly separated and relatively narrow scientific sectors, the model fits well Italian academia, and we test it in a newly constructed dataset which collects the journal publications of all Italian university professors. We find that individual researchers respond to incentives in the manner predicted by the model: more capable researchers respond to increase in the importance of the measurable determinants of promotion and in the competitiveness of the scientific sector by exerting more effort; less able researchers do the opposite. The model predicts a reduction of effort following a new rule introduced in 1999; this prediction finds empirical confirmation in the data.

9.45-9.55 Room change

10.10-12.40 Parallel sessions – 4


• School district consolidation: Scale economies and the optimal scale of education. Fritz Schiltz (KU Leuven) With Kristof de Witte (KU Leuven en TIER Maastricht University)

This paper investigates the relationship between school district size, cost per student and variables measuring the performance of school districts. Based on a comprehensive panel data set covering 1083 school districts, our findings indicate sizeable potential cost savings and performance improvements from consolidation of school districts, especially at the lower tail of the district-size distribution. In order to disentangle scale economies and the optimal scale for each cost type, we apply our cost model to disaggregated cost data, controlling for a wide range of district socioeconomic characteristics. Our



conclusions are supported when repeating the analysis using a production function with teacher absenteeism and school district performance (estimated by a Benefit-of-the-Doubt composite indicator) as outcome measures. Although our application is limited to compulsory education in Flanders, our methodology can be easily extended to other countries or non-profit sectors looking for reduced per unit costs.

• Italian school principals’ managerial behaviors and students’ test scores: an empirical analysis. Mara Soncin (Politecnico di Milano, School of Management) With Tommaso Agasisti (Politecnico di Milano, School of Management) and Patrizia Falzetti (Istituto Nazionale per la Valutazione del Sistema Educativo di Istruzione e Formazione This research investigates the impact of managerial practices implemented by Italian school principals on students’ outcomes. We use micro-data provided by the National Evaluation Committee for Education (INVALSI) for 2013/14 school year. Employing an educational production function, we regress a set of student and school’s characteristics, enriched by information from a questionnaire filled by school principals to estimate student’s score at grade 8 (last year of junior secondary school), also taking into account student’s prior achievement (at grade 6 – first year of junior secondary school). We find that the model well fits for student’s characteristics, while managerial practices tend to have positive effects, but low statistical significance. Stronger associations between management variables and test scores are detected for low-SES schools.

• Private schools and student achievement. Jane Friesen (Simon Fraser University) With Ebrahim Azimi and Simon Woodcock

This paper uses a unique longitudinal, student level data set that includes the population of students enrolled in both private and public primary schools, to estimate private school effects on student achievement. In contrast to most previous evidence on private school effectiveness in advanced economies, we find that private schools on average outperform public schools by between 0.10 and 0.25 standard deviations in both reading and numeracy. These results are robust to a variety of specifications and to reasonable assumptions about selection on unobservable student characteristics. Effect sizes are similar for private “prep” schools, Catholic schools and non-Christian faith schools. Other (non-Catholic) Christian private schools are no more successful than public schools on average with respect to test scores.

• Parents, Schools and Human Capital Differences across Countries.

Marta De Philippis (LSE and Bank of Italy) With Federico Rossi Results from international standardized tests show large cross-country differences in students’ performances. Where do these gaps come from? This paper argues that differences in cultural environments and parental inputs may be of great importance. We show that the school performance of second generation immigrants is closely related with the one of native students in their parents’ countries of origin. This holds true even after accounting for different family background characteristics, schools attended and selection into immigration. We quantify the overall contribution of various parental inputs to the observed cross-country differences in the PISA test performance, and show that they account for



between 12% and 26% of the total variation and for the majority of the gap between East Asia and other regions. This pattern questions whether PISA scores should be interpreted only as a quality measure for a country’s educational system. They actually contain an important intergenerational and cultural component.

• The demand for autonomous schools.

Marco Bertoni (Department of Economics and Management, University of Padova, Italy) With Stephen Gibbons and Olmo Silva (Department of Geography and Environment, Centre for Economic Performance and Spatial Economics Research Centre, London School of Economics; and IZA, Bonn.

Advocates of reforms aimed at increasing school autonomy and parental choice in education argue that market-based incentives have the potential to improve standards. However, one potential concern with such settings is that students with different backgrounds might end up segregated in different schools if parental preferences over school attributes are heterogeneous. In this paper, we study parents’ demand for autonomous schooling using English administrative data about school applications for three cohorts of children choosing secondary schools. We focus on parental preference for public schools that convert to ‘academy’ status – that is, state-funded schools that change their institutional settings to gain independence from the control of the local education authority in terms of taught curriculum, length of the school day and staff management practices (among other things). In order to partial out the effect of time-invariant student and school unobservables, we control for school and pupil fixed-effects. We also conduct a number of checks that deal with time-varying confounders at the school, neighbourhood and student level. Our findings reveal that on average parents express higher demand for autonomous school, although this effect is not very sizeable. However, there is substantial heterogeneity in parental preferences for autonomous education. Parents of high-achieving pupils and better-off households show much stronger preferences for autonomous schools than families with more disadvantaged backgrounds. This finding highlights a potential equity/efficiency trade-off of education systems centred on parental choice and autonomous schooling.


• Where are the resilient schools and what makes the school resilient. Jean Hindriks (Université Catholique de Louvain) With Mattéo Godin (CRED, Université de Namur) In this paper we develop a measure of rank mobility between socio-economic background and academic attainments. We then use PISA 2012 international test scores in Math to estimate both the resilience and the social mobility of the different school systems in the OECD countries. Using the social mobility concept, we revisit the education equity and efficiency trade-off. We show that the social mobility is positively correlated to the education performance (i.e., average test score) and negatively correlated to the education inequality (i.e., variance of test scores). We also show that social mobility is inversely related to the segregation dissimilarity index. In the second part of the paper, we estimate what accounts at the school level for the differences in social mobility using a cross-country regression with random effects to control for country specific aspects. The explanatory variables are related to the school composition (average socio-economic index, social mix, ethnical background, truancy,…) , the school policy (grade repetition, ability grouping, teaching differentiation, school practices for behavioral problems and low achievers, extra-curricular activities, soft skill teaching, academics standards ) and the school autonomy and accountability (budget autonomy, human resource policy, accountability policy…).



• Short- and long-term effects of elite high schools. Mika Kortelainen (VATT Institute for Economic Research) With Ohto Kanninnen

Short- and long-term effects of elite high schools The question of precisely how much students’ skills and educational outcomes are affected by their schools and peers is fundamental for improving education policies, and the answer remains to be fully understood. We are interested in both the effect of school on a single student's results and the distributional effects of student allocation on all the students' results and especially on long-term labor market outcomes. One of the fundamental reasons for our relative lack of knowledge in this field especially regarding long-term outcomes is that educational systems and the process of education are complex systems that rarely allow experimentation. In addition, the fact that different schooling systems in different countries and regions have distinct properties makes comparisons difficult and means that policies need to be assessed separately for each individual country. In this paper we study the causal effects of elite high schools in Finland on educational and long-term labor market outcomes exploiting the structure of the Finnish high school application system. Importantly, Finland has a nation-wide application system that matches the preferences of the applicants and preset student quotas for the schools in a centralized manner. The selection itself is based on the announced preferences and students' grade point averages (GPA) according to the announced available slots in the high schools. This system with school- and year-specific entrance thresholds offers a quasi-experimental regression discontinuity design (RDD), which we exploit using data on Finnish high school students for multiple years (1987-2010). Following studies from other countries, we will concentrate on highly selective elite schools in larger cities that dominate school league tables based on raw exam success. We will investigate both the effect of high schools on educational achievements in the standardized matriculation exams and in tertiary education as well as long-term effects on labor market outcomes (e.g. employment status and earnings).

• Estimating the impact of health on NEET status Gurleen Popli (Department of Economics, University of Sheffield) With Daniel Gladwell and Aki Tsuchiya This paper uses a dynamic Structural Equation Model of ability formation to investigate the determinants of NEET (not in education, employment or training) status in adolescents, with special focus on health. The model addresses the issue of measurement error in estimating ability and mental health; and explores the determinants of ability and NEET status through time. The analysis finds that ability remains the key predictor of NEET status; and while general health plays an important role in the formation of ability for both girls and boys, the impact of mental health differs between the sexes.

• The effect of parental involvement on digital practice for math- and language of

secondary students Carla Haelermans (TIER Maastricht University) With Joris Ghysels (TIER Maastricht University) In this study we analyze the effect of parental involvement in the use of a digital homework practice tool for improving basic math and language skills of students in grade 7 to 9 in the Netherlands, by analyzing the use of an app in which parents can follow their child’s practice behavior in the digital homework tool, using randomized field experiments at the



individual level. The experiment ran one semester and included 3195 students in 3 schools in the Netherlands. After a pretest for math and language in September 2014, students wrote a posttest in January 2015. Half of the parents of these students were allowed to download the app, only half of them really did download, and only 80% of those actually used the app. A parental questionnaire for additional background information generated 27% response, of whom 30% had downloaded the app. Students were supposed to practice an hour per week with the tool, 30 minutes for math and 30 minutes for language. However, data on practice behavior shows that only 80% of students comply, that there is large variation in practice behavior, and that the average is not even near those 30 minutes. There is a large correlation between practice behavior and parental involvement. An IV-analysis to control for non-compliance and amount of practice shows that parental involvement via app-use positively affects practice behavior of 7th grade students, but negatively affects practice behavior of 9th grade students.

• Enrollment and degree completion in higher education without admission standards. Koen Declercq (KU Leuven) Frank Verboven (KU Leuven) The organization of higher education often involves difficult tradeoffs between the objectives of enrollment and degree completion within a reasonable time. On the one hand, governments aim to ensure broad access to a large number of students. On the other hand, they want to allocate resources efficiently and minimize drop-out or delay by matching students to educational programs according to their skills. A recent policy report (OECD, 2012) illustrates the problems in the organization of higher education: up to 62% of today's young adults in OECD countries enter a university-level program, but only 39% are expected to complete it. Among the students who complete a degree, there is a large fraction that incurs substantial delays.


• The Effect of Adaptive versus Static Practicing on Student Learning – Evidence from a Randomized Field Experiment Ilja Cornelisz (TIER UvA) With Chris van Klaveren and Sebastiaan Vonk

This study evaluates the algorithms underlying a static (non-dynamic) and an adaptive (dynamic) version of computerized practicing. Dutch secondary school students (N=1021) are randomly assigned to one version for the duration of one school year. Students using the adaptive version receive more difficult exercises, spend more time practicing, but answer fewer questions correctly. Both groups achieve similar test scores, but students performing sufficiently well prior to the intervention perform slightly worse when working with the adaptive version (SD=-0.08). Adaptive practicing specifically aims at better addressing individual learning needs, but can come at the expense of effective test preparation.

• Fertility Timing and Education.

Sunčica Vujić (University of Antwerp) with Jonathan James (University of Bath)

This paper examines the effect of two education reforms in England and Wales on the timing of fertility. First, we use a feature of the schooling system that led some individuals, due to their month of birth, to being able to leave the school before taking the final high



school exams, and so could leave without having a qualification – the Easter Leaving Rule (ELR). The ELR determined exactly when in the school year people could leave school. Depending on their birthday, children faced one of the two possible leaving dates the end of the Easter term or the end of the summer term. Specifically, those born between the 1st September and the 31st January could leave at Easter while those born between the 1st February and the 31st August had to stay until the end of the summer, enabling them to take end of that year exams, typically held in May and June. After the minimum school leaving age rose to 16 in 1973, late leavers were significantly more likely to obtain academic qualifications. Second, we examine a combination of changes in policies that led to a large expansion in education in the late 1980s and early 1990s, which significantly raised education levels across the whole education distribution, thereby considerably reducing the number of individuals with low education levels in birth cohorts exposed to the expansion. Overall, the proportion of 18 year olds in full time education rose from around 17% in 1985 to over 35% in the late 1990s. Further, the proportion of women with a college degree increased from 13% to 30% from the late 1980s to the early 1990s. We call this period of change the education expansion (EE). We find that neither the exogenous increase in qualifications as a result of the Easter Leaving Rule nor the expansion in post-compulsory schooling led to a reduction in the probability of having a child as a teenager. However, we do find that both sources of variation in education led to delays in having a child. There is no evidence that the mechanism driving these findings is due to an incapacitation effect. Instead the results point to both a direct human capital effect and an improvement in labour market opportunities as a result of holding qualifications.

• Education and skill matching in the European Labour Markets. Some preliminary evaluations from a multilevel analysis. Grazia Graziosi, (University of Trieste) With Laura Chies, (University of Trieste) We are interested in estimating the likelihood on the perception of being over-skilled and both over-skilled and over-qualified, i.e., mixed mismatched, of employees, in relation to what is required to do at their workplace and depending on the level of education. Using data from the first European Skills Survey completed by 48,876 adult employees across all the European Union's 28 Member States, we specify a multilevel logistic model following the general form of the generalized linear mixed model (GLMM) in order to disentangle the differences of both over-skilling and mixed mismatch perception considering two sources of variation: individual and country effects. Preliminary results show that the gender, i.e., being woman, having a semi or elementary education and temporary employment agency contract increase the log-odds of both being over-skilled and mixed mismatched perception. Further, random-effect estimate indicates that the source of variability at the level country plays a non-negligible role in interpreting results.

• Do Rewards Reduce Effort? Evidence From University Exams

Malte Sandner (Leibniz University Hanover) David Kiss & Daniel Schnitzlein (Leibniz University Hanover)

Using data of business administration students, we investigate the effects of bonus points earned during the semester on final exam performance. Students could earn additional final exam points if they passed a simulated exam (called mock exam), which was written



during the semester. Difference-in-differences estimates suggest that receiving the bonus points worsens final exam performance of male students, but did not have any effect on female students.

12.55-14.15 Lunch | Court Yard

14.30-15.15 Keynote speaker: Kristof De Witte (Maastricht University, KU Leuven) | Room ‘Aula Maastricht University’ – Minderbroedersberg 4-6, Maastricht Title: Effectiveness and Efficiency of Educational Innovations Abstract: Despite the fact that we organize education in a very similar way as decades ago, there are daily innovations in education. While most of these innovations are costly, advocates point to their effect on efficiency and effectiveness. The former denotes that outcomes, e.g. educational attainments, can be achieved with fewer resources, or that with the same resources can be obtained. The latter – effectiveness – indicates that the expected outcomes are obtained. This chair continues to this stream of ideas by looking at the efficiency and effectiveness of educational innovations in education. During the inaugural lecture of the chair ‘Effectiveness and Efficiency of Educational Innovations’, I provide an overview of my earlier work. In particular, using insights from earlier papers, I will answer the question how efficiency and effectiveness are intertwined in educational innovations; how resource-driven policies influence educational innovations; how we can measure educational performance in an accurate way; and I discuss the effectiveness of some interventions in early school leaving. Finally, I elaborate some avenues for further research.

15.15-17.00 Drinks

17.00 End of the Workshop

17.00 … For those of you who have some extra free time, a visit to Aachen, Brussels, or Antwerp is definitely worthwhile

second workshop on education economics

Documents