Essays in Honor of M. Hashem Pesaran: Panel Modeling, Micro Applications, and Econometric Methodology: Volume 43B

Many data situations require the consideration of network effects among the cross-sectional units of observation. In this chapter, the authors present a generalized panel model which accounts for two features: (i) network effects present through weighted dependent variables as regressors, exogenous variables, as well as the error components, and (ii) higher-order network effects due to ex ante unknown network decay functions or the presence of multiplex (multi-layer) networks among all of those. The authors outline the model, the basic assumptions, and present simulation results.

The authors provide a way to represent spatial and temporal equilibria in terms of error correction models in a panel setting. This requires potentially two different processes for spatial or network dynamics, both of which can be expressed in terms of spatial weights matrices. The first captures strong cross-sectional dependence, so that a spatial difference, suitably defined, is weakly cross-section dependent (granular) but can be non-stationary. The second is a conventional weights matrix that captures short-run spatio-temporal dynamics as stationary and granular processes. In large samples, cross-section averages serve the first purpose and the authors propose the mean group, common correlated effects estimator together with multiple testing of cross-correlations to provide the short-run spatial weights. The authors apply this model to the 324 local authorities of England, and show that our approach is useful for modeling weak and strong cross-section dependence, together with partial adjustments to two long-run equilibrium relationships and short-run spatio-temporal dynamics. This exercise provides new insights on the (spatial) long-run relationship between house prices and income in the UK.

The authors use a reduced form state-dependent labor participation decision model to illustrate that parameter stability is achieved only if a model properly takes account the observed sample heterogeneity and unobserved sample heterogeneity provided (external) conditions of a model stay constant. Our analysis of the dynamic response path to a health shock using Australian HILDA panel data from 2002 to 2009 shows that experiencing an event by itself can only have a temporary effects. The long-run equilibrium condition is independent of initial conditions or shocks that do not last. In other words, if experiencing an event does not lead to changes in the response parameters such as the real business cycle (Kydland & Prescott, 1977, 1982) or dynamic stochastic general equilibrium model (DSGE, e.g., Sbordone et al., 2010) assumed, policy change may only change the short-run response path. There is no long-term impact for a policy change. On the other hand, if a policy change leads to changes in the decision rules (e.g., the recent US–China trade friction) as the Lucas critique (1976) implies, then there is no other way to evaluate the impact of a policy except to explicitly model how agents respond to the policy change.

Panel data provide the possibilities of estimating individual treatment effects for multiple individuals. Two issues are considered: (1) differences in the estimated individual treatment effects are due to heterogeneity or a chance mechanism? (2) what is the best way to estimate the average treatment effects? Testing and aggregation methods are suggested. Monte Carlo simulations are also conducted to shed light on these two issues. An empirical analysis on the involvement of underground organization in China’s Peer-to-Peer (P2P) activities through the “anti-gang” campaign is also provided.

This chapter analyzes the properties of an alternative least-squares based estimator for linear panel data models with general predetermined regressors. This approach uses backward means of regressors to approximate individual specific fixed effects (FE). The author analyzes sufficient conditions for this estimator to be asymptotically efficient, and argue that, in comparison with the FE estimator, the use of backward means leads to a non-trivial bias-variance tradeoff. The author complements theoretical analysis with an extensive Monte Carlo study, where the author finds that some of the currently available results for restricted AR(1) model cannot be easily generalized, and should be extrapolated with caution.

In the two seminal papers Anderson and Hsiao (1981, 1982), the linear panel regression model without cross-sectional correlation is thoroughly discussed. This uncorrelatedness assumption is now often examined in empirical work, using tests such as those by Pesaran, Ullah, and Yamagata (2008), Hsiao, Pesaran, and Pick (2012), or Pesaran (2015). All these tests in turn improve upon the so-called error-components test suggested in Breusch and Pagan (1980). In this chapter, the author revisits this error-components test and derives its asymptotic distribution under various scenarios: (a) both time-series dimension T and cross-sectional dimension N go to ∞ jointly (Phillips & Moon, 1999); (b) T → ∞ while N is fixed, and (c) N → ∞ while T is fixed. To the best of the author’s knowledge, the results under Scenarios (b) and (c) are new. Moreover, while the distributions under (a) and (b) are normal, that under (c) is not and it is even asymmetric. The critical values under (c) can be simulated. A Monte Carlo experiment is performed and it aims to throw light on the choice among the critical values suggested in the three scenarios, given a T and an N.

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.

Productivity growth in Italy has been persistently anemic and lagged that of the euro area over the period 1999–2015, while the indebtedness of its corporate sector increased. Using the ORBIS firm-level database, this chapter studies the long-term impact of persistent corporate-debt accumulation on the productivity growth of Italian firms, and investigates whether total factor productivity (TFP) growth varies with the level of corporate indebtedness. The authors employ a novel estimation technique proposed by Chudik, Mohaddes, Pesaran, & Raissi (2017) to account for dynamics, bi-directional feedback effects, cross-firm heterogeneity, and cross-sectional dependence arising from unobserved common factors (e.g., oil price shocks, labor and product market frictions, and the stance of the global financial cycle). Filtering out the effects of unobserved common factors and controlling for firm-specific characteristics, the authors find significant negative effects of persistent corporate-debt build-up on firms’ TFP growth on average, and weak evidence of a threshold level of corporate debt, beyond which productivity growth drops off significantly. The results have strong policy implications, for example the design of the tax system should discourage persistent corporate-debt accumulation, and effective and timely frameworks to reduce corporate-debt overhangs are essential.

Building on recent advances in inverse probability weighted identification and estimation of counterfactual distributions, the authors examine the history of wage earnings for women and their potential wage distributions in the United States. These potentials are two counterfactuals, what if women received men’s market “rewards” for their own “skills,” and what if they received the women’s rewards but for men’s characteristics? Using the Current Population Survey data from 1976 to 2013, the authors analyze the entire counterfactual distributions to separate the “structure” and human capital “composition” effect. In contrast to Maasoumi and Wang (2019), the reference outcome in these decompositions is women’s observed earnings distribution, and inverse probability methods are employed, rather than the conditional quantile approaches. The authors provide decision theoretic measures of the distance between two distributions, to complement assessments based on mean, median, or particular quantiles. We assess uniform rankings of alternate distributions by tests of stochastic dominance in order to identify evaluations robust to subjective measures. Traditional moment-based measures severely underestimate the declining trend of the structure effect. Nevertheless, dominance rankings suggest that the structure (“discrimination”?) effect is bigger than human capital characteristics.

The authors provide new evidence in favor of the expectation hypothesis (EH) as a long-run theory of the term structure of interest rates. Using nonparametric techniques first, the authors show that the results of conventional tests that reject EH are strongly affected by the presence of extreme observations – only a handful in the case of longer maturities. The authors then provide a new general methodology that determines the number of outliers causing any theory to fail, and their approach quantifies the extent of this failure.

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.

This chapter extends the work of Baltagi, Bresson, Chaturvedi, and Lacroix (2018) to the popular dynamic panel data model. The authors investigate the robustness of Bayesian panel data models to possible misspecification of the prior distribution. The proposed robust Bayesian approach departs from the standard Bayesian framework in two ways. First, the authors consider the ε-contamination class of prior distributions for the model parameters as well as for the individual effects. Second, both the base elicited priors and the ε-contamination priors use Zellner’s (1986) g-priors for the variance–covariance matrices. The authors propose a general “toolbox” for a wide range of specifications which includes the dynamic panel model with random effects, with cross-correlated effects à la Chamberlain, for the Hausman–Taylor world and for dynamic panel data models with homogeneous/heterogeneous slopes and cross-sectional dependence. Using a Monte Carlo simulation study, the authors compare the finite sample properties of the proposed estimator to those of standard classical estimators. The chapter contributes to the dynamic panel data literature by proposing a general robust Bayesian framework which encompasses the conventional frequentist specifications and their associated estimation methods as special cases.

The authors propose inference methods for endogeneity parameters in linear simultaneous equation models allowing for weak identification and missing instruments. Endogeneity parameters measure the impact of unobserved variables which may be correlated with observed explanatory variables, and play a central role in determining the “bias” associated with endogeneity and measurement errors in structural equations. These results expand, in several ways, the finite-sample theory in Doko Tchatoka and Dufour (2014) for this problem. The latter theory relies on relatively restrictive assumptions, in particular the hypothesis that the reduced form is complete (e.g., contains all the relevant instruments), which is questionable in many practical situations. While the new proposed inference methods retain identification robustness, they also allow the reduced form to be incomplete, for example, due to missing instruments. The authors propose easily applicable inference methods for endogeneity parameters – in particular, two-stage procedures (similar to those in Dufour, 1990). An application to a model of returns to schooling is presented.