Research
Research Papers
(titles subject to change)
Short-term and long-term measures of peer behavior (final version, with Danny Steinbach)
Abstract
Imagine that an individual, who is known as a top performer in an activity, temporarily experiences irregularities so that for certain time periods performance is much lower than expected (and vice versa). The question we pose is: Do the individual’s peers respond to the former (long-term measure) or to the latter (short-term measure)? To answer this question, we exploit unique field panel data on cargo-warehouse agents consolidating freight onto pallets with the help of a forklift. The institutional setting allows for clean identification of both types of social effects. The concept of space is very important as spatial proximity is necessary for the existence of peer effects. We find that agents respond more to their peers’ short-term than their long-term measure of performance: Individuals have a short memory regarding peer behavior. The result implies that changes in team composition are not necessary in order to positively affect outcomes through social interactions.
Presentations:
COPE, Colloquium on Personnel Economics, Augsburg, Germany (plenary session).
EALE, Annual Conference of the European Association for Labour Economists, Lyon, France.
SOLE, Society of Labor Economists, Annual Conference, Toronto, Canada.
Workshop in Social Interactions, Laval University, Quebec, Canada (invited speaker).
AEA, Annual Meeting of the American Economic Association, Philadelphia, U.S.A.
Workshop in Behavioral Economics, IIES, Stockholm University, Stockholm, Sweden.
CRETE, Milos, Hellas.
IPDC, International Panel Data Conference, Thessaloniki, Hellas.
IAAE, Annual Conference of the International Association for Applied Econometrics, Sapporo, Japan.
SOFI Seminar in Labor Economics, Stockholm, Sweden.
SOFI Networks Workshop, Stockholm, Sweden.
Common pitfalls in peer-effects models
Abstract
The majority of the literature in economics does not estimate endogenous social interactions effects (peers' outcome on own outcome) calling upon the reflection problem; instead it resorts to models that estimate only peer effects from exogenous characteristics and considers them as reduced-form parameters that embody the whole social effect. This is not the case. In this paper, first, I use students' performance data from school and a standardized test in compulsory education to estimate endogenous peer effects; the underlying mechanism is verbal interactions among classmates during instruction. Second, and most importantly, I calculate marginal effects for exogenous peer variables from the reduced form which depends nonlinearly on the endogenous-effects parameter and show they can be distinct from coefficients estimated with models using only exogenous peer variables; the latter simply capture omitted-variables bias.
Presentations:
Workshop in Econometrics and Statistics, Southampton, UK (invited speaker).
SOFI Seminar in Labor Economics, Stockholm, Sweden.
Cross-sectional Dependence in Panel Data Models, Trinity College, University of Cambridge, UK.
VfS, German Economic Association Annual Conference.
EALE, Annual Conference of the European Association for Labour Economists, Bonn, Germany.
Seminar at ZEW, Mannheim, Germany (invited speaker).
ACDD 9th Edition, Strasbourg, France.
26th (EC)2 Conference, Theory and Practice of Spatial Econometrics, Edinburgh, UK.
VfS, German Economic Association Annual Conference.
IAAE, Annual Conference of the International Association for Applied Econometrics, Thessaloniki, Hellas.
ETH Zurich, KOF Research Seminar, Zurich, Switzerland (invited speaker).
14th International Workshop on Spatial Econometrics and Statistics, Paris, France.
Crime during the Global Financial Crisis (draft in preparation)
Abstract
I use the global financial crisis in 2008 as an exogenous shock to the German labor market to elucidate the relationship between regional crime and labor market characteristics in the presence of spatial components. The crisis hit export-oriented manufacturing firms located mainly in the west and south as well as their manpower. For the period around the crisis, i.e., 2008-2010, I find that crime such as bodily injury, damage to property and housing burglaries increased in regions with a high proportion of manufacturing and insolvent firms. The effect originates not from unemployed persons as in previous research but from persons employed with reduced working hours (“Kurzarbeit”). The spatial component is important because effects originate not only from within a region but also from neighboring regions explaining diffusion of the crisis in the whole country.
Peer Effects in an Early Tracking System (with Horst Entorf)
Abstract
The German school system applies early tracking: After completing grade 4 around age 10, students are streamed into mainly the Hauptschule leading to vocational training and apprenticeship, the Realschule leading to full-time vocational training, and the Gymnasium leading to university education. First, we provide evidence that students with lower socio-economic status and immigration background are more likely to attend the Hauptschule, so that early tracking impedes educational equality and social mobility. Then, by estimating ability peer effects in 9th-grade classrooms, we show that the performance of outliers as well as minorities have more pronounced effects on students attending the Hauptschule than the Gymnasium. The practice of concentrating low performers and immigrant students in the Hauptschule exacerbates deficiencies through the mechanics of the social multiplier, especially in subjects that depend more on verbal skills.
Slides
Presentations:
International Economic Policy Research Seminar, Goethe University Frankfurt, Germany.
17th Journées Louis-André Gérard-Varet, International Conference in Public Economics, Aix-en-Provence, France.
SOFI Brown Bag Seminar, Stockholm, Sweden.
Seminar at Linnaeus University, Vaxjo, Sweden (invited speaker).
Seminar at IFAU, Uppsala, Sweden (invited speaker).
Other Papers
Migrants at School: Educational Inequality and Social Interaction in the UK and Germany, with Horst Entorf (2009). IZA DP No. 4175
Social Capital and Economic Growth: Evidence from OECD Countries, with Tasneem Zafar (2010). Mimeo, Goethe University. PDF
Other Projects
"Is Pink the New Black? Women at the Corporate Top" with Marie Lalanne (principal investigator). Project quit by Marie Lalanne
Presentations:
SOFI Brown Bag Seminar, Stockholm, Sweden.
Money and Macro Brown Bag Seminar, Goethe University, Frankfurt, Germany.
"Hybrid Data in Household Finance" with Michael Haliassos (principal investigator). Project ended before data collection.