Search results

This chapter develops robust panel estimation in the form of trimmed mean group estimation for potentially heterogenous panel regression models. It trims outlying individuals of which the sample variances of regressors are either extremely small or large. The limiting distribution of the trimmed estimator can be obtained in a similar way to the standard mean group (MG) estimator, provided the random coefficients are conditionally homoskedastic. The authors consider two trimming methods. The first one is based on the order statistic of the sample variance of each regressor. The second one is based on the Mahalanobis depth of the sample variances of regressors. The authors apply them to the MG estimation of the two-way fixed effects model with potentially heterogeneous slope parameters and to the common correlated effects regression, and the authors derive limiting distribution of each estimator. As an empirical illustration, the authors consider the effect of police on property crime rates using the US state-level panel data.

A regression kink design (RKD or RK design) can be used to identify casual effects in settings where the regressor of interest is a kinked function of an assignment variable. In this chapter, we apply an RKD approach to study the effect of unemployment benefits on the duration of joblessness in Austria, and discuss implementation issues that may arise in similar settings, including the use of bandwidth selection algorithms and bias-correction procedures. Although recent developments in nonparametric estimation (Calonico, Cattaneo, & Farrell, 2014; Imbens & Kalyanaraman, 2012) are sometimes interpreted by practitioners as pointing to a default estimation procedure, we show that in any given application different procedures may perform better or worse. In particular, Monte Carlo simulations based on data-generating processes that closely resemble the data from our application show that some asymptotically dominant procedures may actually perform worse than “sub-optimal” alternatives in a given empirical application.

This paper seeks to propose a univariate robust control chart for location and the necessary table of factors for computing the control limits and the central line as an alternative to the Shewhart X¯ control chart.

The proposed method is based on two robust estimators, namely, the sample median, MD, to estimate the process mean, μ, and the median absolute deviation from the sample median, MAD, to estimate the process standard deviation, σ. A numerical example was given and a simulation study was conducted in order to illustrate the performance of the proposed method and compare it with that of the traditional Shewhart X¯ control chart.

The proposed robust X¯_MDMAD control chart gives better performance than the traditional Shewhart X¯ control chart if the underlying distribution of chance causes is non‐normal. It has good properties for heavy‐tailed distribution functions and moderate sample sizes and it compares favorably with the traditional Shewhart X¯ control chart.

The most common statistical process control (SPC) tool is the traditional Shewhart X¯ control chart. The chart is used to monitor the process mean based on the assumption that the underlying distribution of the quality characteristic is normal and there is no major contamination due to outliers. The sample mean, X¯, and the sample standard deviation, S, are the most efficient location and scale estimators for the normal distribution often used to construct the X¯ control chart, but the sample mean, X¯, and the sample standard deviation, S, might not be the best choices when one or both assumptions are not met. Therefore, the need for alternatives to the X¯ control chart comes into play. The literature shows that the sample median, MD, and the median absolute deviation from the sample median, MAD, are indeed more resistant to departures from normality and the presence of outliers.

Educational policies are widely recognized as the means par excellence to equalize opportunities among children with different social and family backgrounds and to promote intergenerational mobility. In this chapter, we focus on the French case and we apply the opportunity equalization criterion proposed by Andreoli, Havnes, and Lefranc (2019) for evaluating the effect of rising compulsory schooling requirements in secondary education. Our results show that such education expansion has a limited redistributive effect on students’ earnings distribution. Nonetheless, we provide evidence of opportunity equalization among groups of students defined by family background circumstances.

The authors develop a novel forecast combination approach based on the order statistics of individual predictability from panel data forecasts. To this end, the authors define the notion of forecast depth, which provides a ranking among different forecasts based on their normalized forecast errors during the training period. The forecast combination is in the form of a depth-weighted trimmed mean. The authors derive the limiting distribution of the depth-weighted forecast combination, based on which the authors can readily construct prediction intervals. Using this novel forecast combination, the authors predict the national level of new COVID-19 cases in the United States and compare it with other approaches including the ensemble forecast from the Centers for Disease Control and Prevention (CDC). The authors find that the depth-weighted forecast combination yields more accurate and robust predictions compared with other popular forecast combinations and reports much narrower prediction intervals.

Researchers in economics and other disciplines are often interested in the causal effect of a binary treatment on outcomes. Econometric methods used to estimate such effects are divided into one of two strands depending on whether they require unconfoundedness (i.e., independence of potential outcomes and treatment assignment conditional on a set of observable covariates). When this assumption holds, researchers now have a wide array of estimation techniques from which to choose. However, very little is known about their performance – both in absolute and relative terms – when measurement error is present. In this study, the performance of several estimators that require unconfoundedness, as well as some that do not, are evaluated in a Monte Carlo study. In all cases, the data-generating process is such that unconfoundedness holds with the ‘real’ data. However, measurement error is then introduced. Specifically, three types of measurement error are considered: (i) errors in treatment assignment, (ii) errors in the outcome, and (iii) errors in the vector of covariates. Recommendations for researchers are provided.

Bagging (bootstrap aggregating) is a smoothing method to improve predictive ability under the presence of parameter estimation uncertainty and model uncertainty. In Lee and Yang (2006), we examined how (equal-weighted and BMA-weighted) bagging works for one-step-ahead binary prediction with an asymmetric cost function for time series, where we considered simple cases with particular choices of a linlin tick loss function and an algorithm to estimate a linear quantile regression model. In the present chapter, we examine how bagging predictors work with different aggregating (averaging) schemes, for multi-step forecast horizons, with a general class of tick loss functions, with different estimation algorithms, for nonlinear quantile regression models, and for different data frequencies. Bagging quantile predictors are constructed via (weighted) averaging over predictors trained on bootstrapped training samples, and bagging binary predictors are conducted via (majority) voting on predictors trained on the bootstrapped training samples. We find that median bagging and trimmed-mean bagging can alleviate the problem of extreme predictors from bootstrap samples and have better performance than equally weighted bagging predictors; that bagging works better at longer forecast horizons; that bagging works well with highly nonlinear quantile regression models (e.g., artificial neural network), and with general tick loss functions. We also find that the performance of bagging may be affected by using different quantile estimation algorithms (in small samples, even if the estimation is consistent) and by using different frequencies of time series data.

We study the statistical properties of Pearson correlation coefficients of Gaussian ranks, and Gaussian rank regressions – ordinary least-squares (OLS) models applied to those ranks. We show that these procedures are fully efficient when the true copula is Gaussian and the margins are non-parametrically estimated, and remain consistent for their population analogs otherwise. We compare them to Spearman and Pearson correlations and their regression counterparts theoretically and in extensive Monte Carlo simulations. Empirical applications to migration and growth across US states, the augmented Solow growth model and momentum and reversal effects in individual stock returns confirm that Gaussian rank procedures are insensitive to outliers.