Marie Thursby, and an anonymous referee. Thanks also to Christian Langer for providing the data on trade flows, to Michael Craw for the US import input-output data, and to Jonathan Putnam for his help in obtaining the technology flow matrix used in the paper.

We propose in this paper a likelihood-based framework for cointegration analysis in panels of a xed number of vector error correction models. Maximum likelihood estimators of the cointegrating vectors are constructed using iterated Generalized Method of Moments estimators. Using these estimators we construct likelihood ratio statistics to test for a common cointegration rank across the individual vector error correction models, both with heterogeneous and homogeneous cointegrating vectors. The corresponding limiting distributions are a summation of the limiting behavior of Johansen (1991) trace statistics. We also incorporate both unrestricted and restricted deterministic components which are either homogeneous or heterogeneous. The proposed framework is applied on a data set of exchange rates and appropriate monetary fundamentals. The test results show strong evidence for the validity of the monetary exchange rate model within a panel of vector error correction models for three major...

In 2006 all ECB publications will feature a motif taken from the €5 banknote. This paper can be downloaded without charge from

In 2007 all ECB publications feature a motif taken from the €20 banknote. This paper can be downloaded without charge from

This paper provides a review of the literature on unit roots and cointegration in panels where the time dimension (T), and the cross section dimension (N) are relatively large. It distinguishes between the …rst generation tests developed on the assumption of the cross section independence, and the second generation tests that allow, in a variety of forms and degrees, the dependence that might prevail across the di¤erent units in the panel. In the analysis of cointegration the hypothesis testing and estimation problems are further complicated by the possibility of cross section cointegration which could arise if the unit roots in the di¤erent cross section units are due to common random walk components.

In 2008 all ECB publications feature a motif taken from the 10 banknote. This paper can be downloaded without charge from

3-Step analysis of

Pedroni, Michael Moore, and Peter Dunne for their comments. This paper has been produced as part of a World Bank research project: Infrastructure and Growth: A Multi-Country Panel Study (RPO 680-89). The findings, interpretations, and conclusions expressed in this paper are entirely those of the authors. They do not necessarily represent We use panel data on a cross section of countries to estimate an aggregate production function including infrastructure capital. We find that the productivity of physical and human capital are close to the levels suggested by microeconomic evidence on their private returns, while electricity generating capacity and transportation networks have roughly the same marginal productivity as capital as a whole. However, telephone networks appear to have a higher marginal productivity than other types of capital. We employ panel data cointegration methods in our estimation that take account of the nonstationary nature of the data, are robust to reverse causation, and allow for different levels of productivity and different short-run business cycle and multiplier relationships across countries

The econometric theory for panel data regressions was largely developed for survey data where N the number of individuals was large and T the number of time periods small. The asymptotic statistical theory was derived by letting N →∞for fixed T. In recent years there has been growing interest in cases, such as sets of countries, regions or industries, where there are fairly long time-series for a large number of groups. These large T, largeN, panel time-series raise a number of issues. First, since time-series tend to be non-stationary, determining the order of integration and cointegration becomes important. Second, since it is possible to estimate a separate regression for each group, it is natural to think of heterogeneous panels where parameters differ over groups. One can then test for parameter homogeneity rather than having to assume it as one is forced to do in small T panels. Third, one needs to determine the asymptotic properties of standard panel estimators when the data are non-stationary. Fourth, thereisanissueofhowtodotheasymptoticanalysisasbothNand T can go to infinity. These issues have generated a large amount of recent econometric work and this paper provides an introductory survey for applied workers.

This paper uses a decomposition of the data into common and idiosyncratic components to develop procedures that test if these components satisfy the null hypothesis of stationarity.