24

The Methods of Growth Econometrics

Steven N. Durlauf, Paul A. Johnson and Jonathan R.W. Temple

Abstract
This chapter provides an overview of current practices in the econometric analysis of economic
growth. We describe some of the main methodologies that have been developed to study growth
as well as some of the major empirical findings with which they are associated. Further, we explore
the relationship between econometric analyses and growth theories. While we argue that there
are a number of respects in which growth econometrics is not adequately integrated with growth
theory, we believe that substantial methodological progress has been made.

24.1 Introduction 1120
24.2 Stylized facts 1123
24.3 Cross-country growth regressions: from theory to empirics 1124
24.3.1 Growth dynamics: basic ideas 1125
24.3.2 Cross-country growth regressions 1126
24.3.3 Levels regressions 1128
24.3.4 Interpreting errors in growth regressions 1129
24.4 Statistical models of the growth process 1130
24.4.1 Specifying explanatory variables in growth regressions 1131
24.4.2 Parameter heterogeneity 1138
24.4.3 Nonlinearity and multiple regimes 1141
24.5 Time series methods, panel data and event studies 1143
24.5.1 Time series approaches 1144
24.5.2 Panel data 1146
24.5.3 The event study approach 1153
24.6 Endogeneity and instrumental variables 1155
24.6.1 Concepts of endogeneity 1155
24.6.2 Exclusion restrictions 1156
24.6.3 Instrumental variables and heterogeneous effects 1158
24.7 Other econometric issues 1159
24.7.1 Outliers 1159
24.7.2 Measurement error 1160
24.7.3 Missing data 1161
24.7.4 Heteroskedasticity 1161
24.7.5 Cross-section error dependence 1162
24.8 Conclusions: the future of growth econometrics 1164

1119