Dynamic Panel Regression Model Research Articles

Global Production Networks and International Inequality: Making a Case for a Meso-Level Turn in Macro-Comparative Sociology

In this article, I extend recent macro-comparative empirical research on the developmental implications of global production networks. I draw from theories of commodity/value chains, global production networks and economic sociology to identify three contending theoretical perspectives for exactly how the developmental returns to network participants should be distributed-cooperation, exploitation and differential gains-and derive testable hypotheses for each. Adding to recent empirical advances for measuring the average network position of firms at the country level, I evaluate these hypotheses by way of dynamic panel regression models of hourly wage rates in the garment and transportation equipment industries. The results suggest that macro-sociological theories linking underdevelopment to the structure of the world-economy, as well as theories of the distribution of the gains from network participation, miss important variation at the industry level. Cooperation provides a poor account of the distribution of the gains from network participation. Instead, both industries appear to distribute the gains from network participation differentially across network participants. However, the extent of this inequality increases, and the garment industry transitions to exploitation, when global production networks become entrenched organizational logics. Variation in the distribution of the returns to network participation is explicable only by accounting for production-network governance as it varies across industries and over time. I conclude by highlighting the analytical utility to macro-comparative sociology of a turn toward the mesa-level of global industries.

Dynamic Linear Panel Regression Models with Interactive Fixed Effects

We analyze linear panel regression models with interactive fixed effects and predetermined regressors, e.g. lagged-dependent variables. The first order asymptotic theory of the least squares (LS) estimator of the regression coefficients is worked out in the limit where both the cross sectional dimension and the number of time periods become large. We find that there are two sources of asymptotic bias of the LS estimator: bias due to correlation or heteroscedasticity of the idiosyncratic error term, and bias due to predetermined (as opposed to strictly exogenous) regressors. A bias corrected least squares estimator is provided. We also present bias corrected versions of the three classical test statistics (Wald, LR and LM test) and show that their asymptotic distribution is a χ2-distribution. Monte Carlo simulations show that the bias correction of the LS estimator and of the test statistics also work well for finite sample sizes.

The Impact of Related Variety on Regional Employment Growth in Finland 1993–2006: High-Tech versus Medium/Low-Tech

This paper investigates the impact of related variety on regional employment growth in Finland between 1993 and 2006 by means of a dynamic panel regression model. We find that related variety in general has no impact on growth. Instead, after separating related variety among low-and-medium-tech sectors from related variety among high-tech sectors, we find that only the latter affects regional growth. Hence, we find evidence that the effect of related variety on regional employment growth is conditioned by the technological intensity of the local sectors involved.

Analysis of interactive fixed effects dynamic linear panel regression with measurement error

This paper studies a simple dynamic linear panel regression model with interactive fixed effects in which the variable of interest is measured with error. To estimate the dynamic coefficient, we consider the least-squares minimum distance (LS-MD) estimation method.

Financial reforms and persistently high bank interest spreads in Bangladesh: Pitfalls in institutional development?

This paper analyzes interest rate spreads and margins in Bangladesh for the period 1990–2008 by applying the Arellano–Bover/Blundell–Bond dynamic panel regression model to a panel of 43 banks. The model has been applied to tackle short-panel bias and endogeneity problems in banking analysis. A high degree of persistency in spreads and margins is observed, which points to inefficiencies of bank management. More specifically, high administrative costs, high non-performing loan ratio, market power, small share of deposits and some macroeconomic factors are found to be the key determinants of persistently high interest rate spreads and margins in Bangladesh. The findings of this study suggest that reforms commenced in the 1990s could not generate adequate competition and efficiency in the financial sector, particularly to drive down the spread in line with the predictions of interest rate literature. This situation in other words indicates pitfalls in institutional development.

Key factors influencing capital structure decision and its speed of adjustment of Thai listed real estate companies

This paper investigates the significant factors influencing capital structure decision of the listed real estate companies and the speed of adjustment towards their target level. The study used homogeneous panel of 39 Thai companies in real estate industry listed in the Stock Exchange of Thailand (SET) during the period 2002 to 2009. The analysis employs multiple linear panel regression models in examining factors influencing capital structure decision, as well as, dynamic panel regression model using one-step and two-step Arellano and Bond GMM estimation methods in determining the speed of adjustment towards target capital structure. The findings indicate that firm leverage is positively related to median industry leverage. Furthermore, firm size and growth opportunities have positive relationship with firm leverage, whereas profitability and leverage are negatively associated. Our results support pecking order theory as higher profitability firms tend to have less debt and firms with higher growth opportunities tend to have greater leverage. Additionally, the study also discovers that real estate companies partially adjust their capital structure towards the target level capital structure only at the rate of 63 percent.

Government policy and industrial investment determinants in Spanish regions

The aim of this paper is to analyse the effect of public policies on private industrial investment in Spanish regions by means of a dynamic panel regression model with data collected from 1980 to 2000. We derive and estimate a Euler investment equation that accounts for the influence of both technological and human capital and public infrastructure, which may affect regional efficiency and, therefore, industrial investment. Our results indicate that investment has been sensitive to public infrastructure, particularly in the 1980s, while the effect of human capital was felt throughout the period.

REDUCING BIAS OF MLE IN A DYNAMIC PANEL MODEL

This paper investigates a simple dynamic linear panel regression model with both fixed effects and time effects. Using “large n and large T” asymptotics, we approximate the distribution of the fixed effect estimator of the autoregressive parameter in the dynamic linear panel model and derive its asymptotic bias. We find that the same higher order bias correction approach proposed by Hahn and Kuersteiner (2002, Econometrica 70, 1639–1659) can be applied to the dynamic linear panel model even when time specific effects are present.We thank Peter Phillips and three anonymous referees for helpful comments. The first author gratefully acknowledges financial support from NSF grant SES-0313651. The second author appreciates the Faculty Development Awards of USC for research support.

Reducing Bias of MLE in a Dynamic Panel Model

This paper investigates a simple dynamic linear panel regression model with both fixed effects and time effects. Using large n and large T asymptotics, we approximate the distribution of the fixed effect estimator of the autoregressive parameter in the dynamic linear panel model and derive its asymptotic bias. We find that the same higher order bias correction approach proposed by Hahn and Kuersteiner (2002) can be applied to the dynamic linear panel model even when time specifc effects are present.

Nonstationary panel data analysis: an overview of some recent developments

This paper overviews some recent developments in panel data asymptotics, concentrating on the nonstationary panel case and gives a new result for models with individual effects. Underlying recent theory are asymptotics for multi-indexed processes in which both indexes may pass to infinity. We review some of the new limit theory that has been developed, show how it can be applied and give a new interpretation of individual effects in nonstationary panel data. Fundamental to the interpretation of much of the asymptotics is the concept of a panel regression coefficient which measures the long run average relation across a section of the panel. This concept is analogous to the statistical interpretation of the coefficient in a classical regression relation. A variety of nonstationary panel data models are discussed and the paper reviews the asymptotic properties of estimators in these various models. Some recent developments in panel unit root tests and stationary dynamic panel regression models are also reviewed.

Spurious Stochastics in a Short Time-Series Panel Data

This paper analyzes the effects of individual-specific size factors in a dynamic panel regression model. Theory and simulation show that an individual-specific size factor, with a fat-tailed distribution or a time- varying property, may cause spurious stochastics. If a pair of panel variables depends on size in some way, then they appear to find a strong relationship, if the size variable is not used in the regression, even if the variables are otherwise independent. Moreover, forecasts based on models that have omitted size-factors are affected seriously by the property of the size-factors. A pooling regression with very short time-series appears to fit well in sample, but forecasts poorly out-of-sample if the neglected indivi- dual-specific size-factor has a fat-tailed distribution.