Galton, Edgeworth, Frisch, and prospects for quantile regression in econometrics

The work of three leading figures in the early history of econometrics is used to motivate some recent developments in the theory and application of quantile regression. We stress not only the robustness advantages of this form of semiparametric statistical method, but also the opportunity to recover a more complete description of the statistical relationship between variables. A recent proposal for a more X-robust form of quantile regression based on maximal depth ideas is described along with an interesting historical antecedent. Finally, the notorious computational burden of median regression, and quantile regression more generally, is addressed. It is argued that recent developments in interior point methods for linear programming together with some new preprocessing ideas make it possible to compute quantile regressions as quickly as least-squares regressions throughout the entire range of problem sizes encountered in econometrics.

[1]  Pin T. Ng,et al.  Quantile smoothing splines , 1994 .

[2]  J. Simonoff.  Multivariate Density Estimation , 1996 .

[3]  C. Müller,et al.  Breakdown points and variation exponents of robust $M$-estimators in linear models , 1999 .

[4]  D. Donoho,et al.  Breakdown Properties of Location Estimates Based on Halfspace Depth and Projected Outlyingness , 1992 .

[5]  S. Stigler.  Francis Ysidro Edgeworth, Statistician , 1978 .

[6]  Roger Koenker,et al.  Tests of linear hypotheses based on regression rank scores , 1993 .

[7]  Sanjay Mehrotra,et al.  On the Implementation of a Primal-Dual Interior Point Method , 1992, SIAM J. Optim..

[8]  Frisch.  La résolution des problèmes de programme linéaire par la méthode du potentiel logarithmique , 1956 .

[9]  J. Tukey.  A survey of sampling from contaminated distributions , 1960 .

[10]  P. Laplace.  Théorie analytique des probabilités , 1995 .

[11]  A. Hald.  A history of mathematical statistics from 1750 to 1930 , 1998 .

[12]  S. Portnoy,et al.  Asymptotics Of The Deepest Line , 1998 .

[13]  F. Y. Edgeworth.  L. The method of least squares , 1883 .

[14]  C. Gutenbrunner,et al.  Regression Rank Scores and Regression Quantiles , 1992 .

[15]  Ian Barrodale,et al.  Algorithm 478: Solution of an Overdetermined System of Equations in the l1 Norm [F4] , 1974, Commun. ACM.

[16]  J. Tukey.  Mathematics and the Picturing of Data , 1975 .

[17]  T. Wet,et al.  Mallows-Type Bounded-Influence-Regression Trimmed Means , 1988 .

[18]  Peter Rousseeuw,et al.  Computing location depth and regression depth in higher dimensions , 1998, Stat. Comput..

[19]  F. Y. Edgeworth.  XXII. On a new method of reducing observations relating to several quantities , 1888 .

[20]  Stephen M Stigler,et al.  Regression towards the mean, historically considered , 1997, Statistical methods in medical research.

[21]  Regina Y. Liu,et al.  Regression depth. Commentaries. Rejoinder , 1999 .

[22]  Ragnar Frisch,et al.  Foundations of modern econometrics : the selected essays of Ragnar Frisch , 1995 .

[23]  Clóvis C. Gonzaga,et al.  Path-Following Methods for Linear Programming , 1992, SIAM Rev..

[24]  Anthony V. Fiacco,et al.  Nonlinear programming;: Sequential unconstrained minimization techniques , 1968 .

[25]  F. E.,et al.  Natural Inheritance , 1889, Nature.

[26]  R. Koenker,et al.  Asymptotic Theory of Least Absolute Error Regression , 1978 .

[27]  Pin T. Ng,et al.  COBS: qualitatively constrained smoothing via linear programming , 1999, Comput. Stat..

[28]  Michael A. Saunders,et al.  On projected newton barrier methods for linear programming and an equivalence to Karmarkar’s projective method , 1986, Math. Program..

[29]  CXIII.On the use of medians for reducing observations relating to several quantities , 1923 .

[30]  J. Powell,et al.  Censored regression quantiles , 1986 .

[31]  D. Freedman,et al.  A solution to the ecological inference problem , 1997 .

[32]  R. Koenker,et al.  The Gaussian hare and the Laplacian tortoise: computability of squared-error versus absolute-error estimators , 1997 .

[33]  F. Galton.  Regression Towards Mediocrity in Hereditary Stature. , 1886 .

[34]  Douglas Rivers.  A Solution to the Ecological Inference Problem: Reconstructing Individual Behavior from Aggregate Data . By Gary King. Princeton, NJ: Princeton University Press, 1997. 342p. $55.00 cloth, $16.95 paper. , 1998 .

[35]  Narendra Karmarkar,et al.  A new polynomial-time algorithm for linear programming , 1984, Comb..

[36]  James L. Powell,et al.  Efficient Estimation of Linear and Type I Censored Regression Models Under Conditional Quantile Restrictions , 1990, Econometric Theory.

[37]  Stephen J. Wright.  Primal-Dual Interior-Point Methods , 1997, Other Titles in Applied Mathematics.

[38]  W. Härdle,et al.  A note on prediction via estimation of the conditional mode function , 1986 .

[39]  John W. Tukey,et al.  Data Analysis and Regression: A Second Course in Statistics , 1977 .

[40]  O. Sheynin,et al.  R. J. Boscovich's work on probability , 1973, Archive for History of Exact Sciences.

[41]  Rob J Hyndman,et al.  Estimating and Visualizing Conditional Densities , 1996 .

[42]  A. Shiryayev.  The Method of The Median in The Theory of Errors , 1992 .

[43]  H. Theil.  A Rank-Invariant Method of Linear and Polynomial Regression Analysis , 1992 .

[44]  Margaret H. Wright,et al.  Interior methods for constrained optimization , 1992, Acta Numerica.

[45]  Philip E. Gill,et al.  Numerical Linear Algebra and Optimization , 1991 .

[46]  Xuming He.  TAIL BEHAVIOR OF REGRESSION ESTIMATORS AND THEIR BREAKDOWN POINTS , 1990 .

[47]  C. Manski.  Semiparametric analysis of discrete response: Asymptotic properties of the maximum score estimator , 1985 .

[48]  Ronald L. Rivest,et al.  Expected time bounds for selection , 1975, Commun. ACM.