Prof. Dr. Alois Kneip

E-mail: akneip(at)
Phone: +49 228 73 9263
Fax: +49 228 735050
Institute: Department of Economics
Research Area: Research Area H (Leader)
Date of birth: 12.May 1956
Mathscinet-Number: 249575

Academic Career


Dr. rer. nat., University of Heidelberg

1988 - 1990

Postdoc, ZI Mannheim

1990 - 1994

Postdoc, Institute of Economics, University of Bonn


Habilitation (Economics), University of Bonn

1994 - 2000

Professor, Université Catholique de Louvain, Belgium

2000 - 2005

Professor (C4), University of Mainz

Since 2005

Professor (W3), University of Bonn

Research Profile

A substantial part of my research focuses on nonparametric statistics and functional data analysis. The scientific approach incorporates the development of new methods, a study of resulting theoretical properties, and real data analysis. Important applications in economics are the study of the development of income distributions, Engel curves, and implied volatility surfaces over time or under different conditions. Biomedical applications include the analysis of human growth curves or gene expression data. Major recent contributions to this area are given in [6], [3] and [5]. A second line of my research aims at quantifying individual heterogeneity in economic panel data. In [2] it has been shown that nonparametric smoothing procedures may serve as a tool to improve efficiency of estimation of unknown factors in factor models. A third line of research considers the econometric analysis of production efficiencies based on frontier models. In a number of papers (e.g. [7], [8], [9] and [10]) we have developed a non-standard, general methodolgy for statistical inference of DEA and FDH estimators.

In the coming years I plan to study some important open problems in functional data analysis. New techniques have to be developed for the analysis of dependent data and, in particular, for time series of functions, which are important for analyzing financial data. Conventional methods for smooth functions have to be adapted in the context of non-smooth economic processes.
A major challenge is to overcome the limitations of standard Hilbert space analysis. Some important applications lead to samples of density functions which lie a nonlinear subspace of L^2. Experiments in biomedicine frequently yield functional data possessing a common structure in terms of typical successions of peaks and valleys. The challenge then consists in identifying low-dimensional, nonlinear manifolds characterizing the data.
Crucial theoretical and methodological questions will have to be resolved. In the context of panel data I plan to consider problems of time varying and individually heterogeneous regression coefficients.

Research Projects and Activities

Research Project H
Principal Investigator, since 2006

Research Area Econometrics/Statistics of the Bonn Graduate School of Economics
Principal Investigator, since 2006

Contribution to Research Areas

Research Area H
My major focus has been the development of statistical/econometrical methods for dealing with heterogeneous populations and corresponding aggregation problems. In [1], we analyze budget elasticities of demand. It is shown that there are significant differences between properly defined aggregate elasticities and mean individual elasticities on the micro level. A rigorous microeconomic theory is developed and the empirical analysis is based on nonparametric statistical methods. When dealing with heterogeneous populations, a major task is to model the time evolutions of the distributions of important characteristics, or of their corresponding density and regression functions. In the last decades functional data analysis has been a very active field of international statistical research. The aim is to develop methods for analyzing data representing samples or time series of related functions. In [2], we combine econometric factor models and methods of functional data analysis in order to analyze heterogeneity in time trends. Functional regression problems are studied [3], while in [4], we present a new approach to the so-called registration problem. Functional principal components analysis is used in [5] to analyze implied volatility surfaces of European options.

Selected Publications

[1] Michal Paluch, Alois Kneip, Werner Hildenbrand
Individual versus aggregate income elasticities for heterogeneous populations
J. Appl. Econometrics , 27: (5): 847--869
DOI: 10.1002/jae.1237
[2] Alois Kneip, Robin C. Sickles, Wonho Song
A new panel data treatment for heterogeneity in time trends
Econometric Theory , 28: (3): 590--628
DOI: 10.1017/S026646661100034X
[3] Christophe Crambes, Alois Kneip, Pascal Sarda
Smoothing splines estimators for functional linear regression
Ann. Statist. , 37: (1): 35--72
DOI: 10.1214/07-AOS563
[4] Alois Kneip, James O. Ramsay
Combining registration and fitting for functional models
J. Amer. Statist. Assoc. , 103: (483): 1155--1165
DOI: 10.1198/016214508000000517
[5] Michal Benko, Wolfgang Härdle, Alois Kneip
Common functional principal components
Ann. Statist. , 37: (1): 1--34
DOI: 10.1214/07-AOS516
[6] Alois Kneip, Dominik Poss, Pascal Sarda
Functional linear regression with points of impact
Ann. Statist. , 44: (1): 1--30
DOI: 10.1214/15-AOS1323
[7] Alois Kneip, Léopold Simar, Paul W. Wilson
Testing hypotheses in nonparametric models of production
J. Bus. Econom. Statist. , 34: (3): 435--456
DOI: 10.1080/07350015.2015.1049747
[8] Alois Kneip, Léopold Simar, Ingrid Van Keilegom
Frontier estimation in the presence of measurement error with unknown variance
J. Econometrics , 184: (2): 379--393
DOI: 10.1016/j.jeconom.2014.09.012
[9] Alois Kneip, Léopold Simar, Paul W. Wilson
When bias kills the variance: central limit theorems for DEA and FDH efficiency scores
Econometric Theory , 31: (2): 394--422
DOI: 10.1017/S0266466614000413
[10] A. Kneip, L. Simar, P. Wilson
A computationally efficient, consistent bootstrap for inference with non-parametric DEA estimators
to appear
Computational Economics
[11] Alois Kneip, Pascal Sarda
Factor models and variable selection in high-dimensional regression analysis
Ann. Statist.
, 39: (5): 2410--2447
DOI: 10.1214/11-AOS905
[12] Alois Kneip, Klaus J. Utikal
Inference for density families using functional principal component analysis
With comments and a rejoinder by the authors
J. Amer. Statist. Assoc. , 96: (454): 519--542
DOI: 10.1198/016214501753168235
[13] Alois Kneip
Ordered linear smoothers
Ann. Statist. , 22: (2): 835--866
DOI: 10.1214/aos/1176325498
[14] Alois Kneip, Theo Gasser
Statistical tools to analyze data representing a sample of curves
Ann. Statist. , 20: (3): 1266--1305
DOI: 10.1214/aos/1176348769

Publication List


• Computational Statistics (2005 – 2010)
• Annals of Statistics (2008 - 2012)
• Bernoulli (2008 – 2010)



Elected Member, International Statistical Institute

Selected Invited Lectures


Workshop on Knowledge Extraction and Modeling, Capri, Italy


ENSAI workshop on Nonparametric Statistics, Rennes, France


COMPSTAT, Rome, Italy


BIRS workshop on Statistics at the Frontiers of Science, Banff, AB, Canada


4èmes Journées STAPH, Grenoble, France


Workshop on Semiparametric and Nonparametric Methods in Econometrics, Oberwolfach


ISI 2007, Lisbon, Portugal


WIAS workshop on Sparsity and Inverse Problems in Statistical Theory and Econometrics, Berlin


1st International Workshop on Functional and Operational Statistics, Toulouse, France


BIRS workshop on Semiparametric and Nonparametric Methods in Econometrics, Banff, AB, Canada


S. Co. 2009, Milan, Italy


BIRS workshop on Functional Data Analysis: Future Directions, Banff, AB, Canada


17th Meeting of AiOs in Stochastics, Hilversum, Netherlands


73rd Annual Meeting of the Institute of Mathematical Statistics, Gothenburg, Sweden


SAMSI workshop on Analysis of Object Data, Research Triangle, NC, USA



Offer of a chair in Statistics, University of Dortmund

Selected PhD students

Florence Nicole (2002): “Registration and Functional Data Analysis”,
now Associate Professor, École Nationale de l'Aviation Civile, Toulouse, France

Supervised Theses

  • Master theses: 7
  • Diplom theses: 35
  • PhD theses: 5
Download Profile