No full-text available
To read the full-text of this research,
you can request a copy directly from the author.
Tests de causalité sur données de panel : une application à l'étude de la causalité entre l'investissement et la croissance
Abstract
Causaliy Testing in Panel Data with an Application to the Question of Investment and Growth
by Diana Weinhold
This paper develops an approach for panel causality analysis which allows for flexibility in the causal relationship from unity to unity, allowing estimation of distribution of causality in a possibly heterogeneous panel. Two possible estimation techniques for "mixed fixed and random" coefficients model are explored. These methods both allow the coefficients of an orthogonalized causal variable to vary randomly while avoiding many of the problems associated with random coefficients and dynamics in panels. Using simulations to develop a method for interpreting the estimated distributions, it is then possible to predict the probabilities of causality associated with the panel.
This approach is then applied to the question of causality between investment and growth in a panel of countries. The paper find that there is great instability and feedback among countries in this relationship. Therefore, it is clear that the restricting assumptions of traditional pooling models can be inappropriate, especially in simple models. Because the analysis signals this problem, the paper's results further underline the need for the interpretive flexibility that this proposed causality test imparts.
... This idea is based on the relative predictive power of models, and is therefore intrinsically part of a time series approach. The preconditions for this test are based on the idea that if one variable X causes another variable Y, then adding lags (Weinhold, 1996) on X in a regression of Y on its own lagged values and other explanatory variables should significantly improve the predictive power of the model (Weinhold, 1996;Aye et al., 2017). The Granger (1969) causality test has been progressively extended to panel data with the work of Society (2011). ...
... This idea is based on the relative predictive power of models, and is therefore intrinsically part of a time series approach. The preconditions for this test are based on the idea that if one variable X causes another variable Y, then adding lags (Weinhold, 1996) on X in a regression of Y on its own lagged values and other explanatory variables should significantly improve the predictive power of the model (Weinhold, 1996;Aye et al., 2017). The Granger (1969) causality test has been progressively extended to panel data with the work of Society (2011). ...
Mobilization of adequate tax revenues remains a key objective for African countries in the implementation of
development policies. The objective of this study is to examine the effect of internet access and cell phone access on tax
revenues in West African Economic and Monetary Union (WAEMU) countries, from 2000 to 2019. The data used are from
the Central Bank of West African States (BCEAO) and World Bank database. The methodology used is based on the
Granger causality test procedure and the fixed effect panel model. The results show that there is a causal relationship
between cell phone access, internet access and tax revenues. The results of the estimates show that internet access and cell
phone access have positive and significant effects on tax revenues, and a 10% increase in internet access increases tax
revenues by 1.23%, and a 10% increase in mobile phone access increases tax revenues by 1.11%. The recommendations are
that WAEMU countries should invest more in the digitalization of the economy and the tax system, improve access to good
quality internet and cell phone access, and promote the use of digital technology in resource mobilization processes.
... This idea is based on the relative predictive power of models, and is therefore intrinsically part of a time series approach. The prerequisites for this test are based on the idea that if variable X causes another variable Y, then adding lags (Weinhold, 1996) on X in a regression of Y on its own lagged values and other explanatory variables should significantly improve the predictive power of the model (Weinhold, 1996;Aye et al., 2017). The causality test of Granger (1969) was gradually extended to panel data with the work of Society (2011). ...
... This idea is based on the relative predictive power of models, and is therefore intrinsically part of a time series approach. The prerequisites for this test are based on the idea that if variable X causes another variable Y, then adding lags (Weinhold, 1996) on X in a regression of Y on its own lagged values and other explanatory variables should significantly improve the predictive power of the model (Weinhold, 1996;Aye et al., 2017). The causality test of Granger (1969) was gradually extended to panel data with the work of Society (2011). ...
Despite the established roles of governance and ICT in development, the literature on good governance has not yet given the link between quality official statistics (quality information) needed for good decision making and good governance a research interest worth examining. The objective of this study is to examine how ICT development complements the development of official statistics in influencing governance outcomes in 33 African countries for the period 2010-2019. Governance outcomes are measured by the Mo Ibrahim Foundation's Governance Index. The ICT Development Index is constructed using the principal component analysis method while the Statistical Capacity Index is derived from the World Bank database. The regression model used is the generalized method of moments (GMM). The results show that there is a causality between the capacity of African statistical offices to produce and make available the data and information needed for good decision-making and governance on the one hand, and a causality from ICT development to statistical capacity and from ICT development to governance on the other. The estimation results show that the net effect on governance is positive, significant and large when the development of official statistics is accompanied by the development of ICT. Overall, the results suggest that investments should be made in both ICT development and official statistics production for good decision making, better governance and socioeconomic development in Africa.
... . Contrary to Weinhold (1996) and Nair-Reichert and Weinhold (2001), we assume that the autoregressive coefficients γ (k) and the regression coefficients slopes β ...
... The results of our simulations are reported on table (1). For each time dimension T, we compare the critical value of the theoretical centered F 1,n 1 , denoted f α (1, n 1 ) , to the empirical critical value get from 5000 replications of model (23). We can observe that, for small T, the empirical critical values are higher than the theoretical ones, due to the noncentrality parameter of the true distribution of F hnc . ...
This paper proposes an extension of the Granger (1969) causality definition to panel data models with fixed coefficients. Given the het-erogeneity of the data generating process, four definitions of causality relationships are proposed. A procedure of nested tests is then defined and a particular attention is done to finite sample properties of these tests. Finally, an application to the link between financial deepening and economic growth is proposed with a panel of sixteen sub-Saharan countries over the period 1967-1998.
... We used the Granger test to examine the existence of a causal relationship between the explanatory variables and intra-African trade openness. In a bivariate framework, one variable causes a second in the Granger sense if the prediction of the second variable improves when the lagged variables of the first variable are taken into account (Granger, 1969;Weinhold, 1996;Aye et al., 2017). We use the Granger causality procedure for panel data proposed by Dumitrescu and Hurlin (2012). ...
The literature seems to focus on tariffs as the main factor limiting trade between African countries. This study highlights other factors that affect trade in Africa. It examines the causality between intra-African trade openness and the transport infrastructure development index, industrial development index, information and communication technologies (ICT) development, the business and competition regulation index, and the Customs duty rate. The panel causality test of Granger (1969) and Dumitrescu and Hurlin (2012) using to data from 36 African countries over the period 2010-2019. The results of the Granger causality test reveal that there is a causal relationship between internet access, mobile phone access, customs duty rate, transport infrastructure development, business and competition regulation and intra-African trade openness. There is a bidirectional causality between intra-African trade openness and the development of the industrial sector on the one hand, and a bidirectional causality between the development of transport infrastructure and the development of the industrial sector on the other. The results suggest African countries should put in place policies that focus on transport infrastructure development and industrial development. African governments should also strengthen the development of communication infrastructure (access to the Internet and mobile telephony) and intensify exemptions from customs duties on imports of equipment acquired in the context of industrial development and the construction of transport infrastructure.
... Besides, we allow the autoregressive parameters γ i (k) and the regression coefficients slopes β i (k) to differ across groups. However, contrary to Weinhold (1996) and Nair-Reichert and Weinhold (2001), parameters γ i (k) and β i (k) are constant in time. It is important to note that our model is not a random coefficient model as in Swamy (1970), it is a fixed coefficient model with fixed individual effects (Dumitrescu & Hurlin, 2012, p. 1451). ...
... In equation (13) are assumed to vary across groups. However, lag parameters and slope coefficients are supposed to be fixed in contrast to Weinhold (1996) and Nair-Reichert and Weinhold (2001). K denotes lag orders and is supposed to be identical for all individuals in balanced panels. ...
... By means of Monte Carlo simulations, we show that the size and power of these tests are seriously deteriorated when the linearity assumption breaks down, when the data generating process is heterogeneous across the cross-section units, when there are structural breaks or extreme observations in some of the cross-section units, when the causal dependence appears in the conditional variance or when it 1 Although this paper focuses on economic applications, there has been recently a growing interest in the use of Granger causality to identify causal interactions in several other fields such as neuroscience and neuroimaging (Seth, Barrett, and Barnett, 2015). 2 Hurlin and Venet (2001) assume homogeneous autoregressive coefficients but they allow for heterogeneity in the regression coefficients slopes. Weinhold (1996) and Nair-Reichert and Weinhold (2001) have considered testing for causality in a panel using random coefficient models. involves qualitative data. ...
In this study, we use multiple-unit symbolic dynamics and transfer entropy to develop a non-parametric Granger causality test procedure for longitudinal data. Monte Carlo simulations show that our test exhibits the correct size and a high power in situations where linear panel data causality tests fail, such as (1) when the linearity assumption does not hold, (2) when the data generating process is heterogeneous across the cross-section units or presents structural breaks, (3) when there are extreme observations in some of the cross-section units, (4) when the process exhibits causal dependence on the conditional variance, or (5) when the analysis involves qualitative data. We illustrate the usefulness of our proposed procedure by analyzing the dynamic causal relationships between public expenditure and GDP, between firm productivity and firm size in US manufacturing sectors, and among sovereign credit ratings, growth, and interest rates.
... Holtz-Eakin et al. (1985) et Weinhold (1996) Notre travail prend également en considération l'approche de Konya (2006) 6 fondée sur un modèle stationnaire pour éliminer les effets de non stationnarité, néfaste à la normalité supposée dans cette approche. L'approche de Konya (2006) n'exige pas, quant à elle, la stationnarité pour effectuer le test de causalité, puisqu'elle est basée sur un bootstrap qui simule la distribution empirique de la statistique considérée. ...
Cet article s’ intéresse au concept de causalité en panel. Les approches de Konya (2006) et de Dumitrescu et Hurlin (2011) ont été retenues pour tester la (non) causalité au sens de Granger (1969) entre le taux de change réel et la croissance économique dans les pays en développement. Nous avons choisi un panel de trois groupes de pays (BRICS 1, l’organisation de coopération de Shanghai (OCS) 2, et les Tigres asiatiques 3) et utilisé le flux d’investissements directs comme variable de contrôle. Les résultats indiquent que le test de Konya (2006) rejette moins l’hypothèse de non causalité que celui de Dumitrescu et Hurlin (2011). L’analyse montre que la sous-évaluation des monnaies des pays du panel (notamment les pays de l’organisation de coopération de Shanghai (OCS)) a causé une baisse de la croissance économique sur la période 1990-2012. Le flux d’investissements directs joue un rôle crucial pour stimuler la causalité entre la croissance économique et le taux de change réel.
... En date du ralentissement économique de la Jordanie qui est également la date de rupture de la performance mesurée par le Q de Tobin, cette variable prend la valeur 1 à partir de l'année 2008. l'estimation de modèles dynamiques à coefficients aléatoires (Weinhold, 1996) (Hendry 1989). ...
... Bien que divisés par des frontières internationales, le Nord de la RCA, l'Est du Tchad et l'Ouest du Soudan sont étroitement liés socialement, avec de nombreux groupes ethniques communs aux deux derniers pays (Giroux et al., 2009 ;De Maio, 2010 (Weinhold, 1996 ;Nair-Reichert et Weinhold, 2001). ...
Cette thèse présente un ensemble de travaux de recherche en économie politique qui s’inscrivent dans le champ contemporain de la littérature sur la malédiction des ressources naturelles dans les pays en développement. Elle s’intéresse spécifiquement aux liens entre les rentes des ressources naturelles et les institutions politiques et les conflits, puis entre les rentes et la pauvreté sous ses différentes formes. Son champ géographique se situe à plusieurs niveaux, international pour un groupe de pays en développement, monographique et infranational pour le Tchad et ses régions. La première partie, en s’inspirant des théories des changements institutionnels, teste l’hypothèse que les rentes issues de différentes ressources naturelles ont des effets différenciés sur la qualité des institutions dans les pays en développement. Par la suite, elle teste l’hypothèse d’un « système de conflit régional » dans le Triangle République Centrafricaine, Tchad et Soudan. La deuxième partie évalue le dispositif « original » de gestion des revenus pétroliers sur différentes dimensions de la pauvreté au Tchad sur des données d’enquêtes de terrain. Le premier chapitre de cette partie teste l’hypothèse d’un biais urbain dans l’allocation de temps des enfants dans un contexte pétrolier au Tchad. Alors que le second est dédié à l’évaluation de l’impact de l’exploitation pétrolière sur le profil de la pauvreté dans la région productrice en utilisant les méthodes quasi-expérimentales d’évaluation d’impact.
... Dumitrescu and Hurlin [53] use the following model in order to test the panel Granger causality [53]: parameters are assumed to be constant in time contrary to Weinhold [65] and Nair-Reichert and Weinhold [66]. The model is established as a fixed coefficient model with fixed individual effects contrary to Swamy [67] who adopts a random coefficient model. ...
This study investigates the causal relationship between energy consumption, urbanization and economic growth using Dumitrescu-Hurlin panel Granger causality test for the period 1971-2014 in New Emerging-Market Countries( Colombia, India, Indonesia, Kenya, Malaysia, and Mexico). Bivariate and trivariate panel Granger causality analyses which detect joint causality effect from two series to another are applied. According to the bivariate analysis, there is panel Granger causality from economic growth to energy consumption, and from urbanization to energy consumption and economic growth. According to the trivariate analysis, there is panel Granger causality from energy consumption and urbanization to economic growth, from economic growth and urbanization to energy consumption, and from energy consumption and economic growth to urbanization.
... Besides, we allow the autoregressive parameters γ i (k) and the regression coefficients slopes β i (k) to differ across groups. However, contrary to Weinhold (1996) and Nair-Reichert and Weinhold (2001), parameters γ i (k) and β i (k) are constant in time. It is important to note that our model is not a random coefficient model as in Swamy (1970), it is a fixed coefficient model with fixed individual effects (Dumitrescu & Hurlin, 2012, p. 1451). ...
Econometric modelling on tourist arrivals has concentrated mostly on forecasting. We abstract and assess the impact on arrivals to Korea from China, Hong Kong, Indonesia, Japan, Taiwan, Singapore and Malaysia using a number of macroeconomics variables (unemployment, inflation, GDP and exchange rate) sentiment, mood and most importantly ‘Hallyu’ (Korean cultural wave) which is quantified for the very first time. Effectively, we test how each variable ‘measures up’ against all other variables. Our analysis is twofold. Firstly, we use impulse response functions to assess the risk, magnitude and duration that each shock has on short-term arrivals from each country separately. Secondly, we concentrate on systemic effects and use causality tests to examine the effect of each variable on arrivals for all countries simultaneously. We find that the effect of our variables is not uniform but most importantly that ‘Hallyu’ cannot be ignored.
... A more or less similar procedure is applied by Hsiao [24], Holtz-Eakin et al., [23] Hsiao [25], Weinhold [26], Weinhold [27], Nair-Reichart and Weinhold [28], and Choe [29]. The second approach proposed by Hurlin and Venet [30], Hurlin [31], Hurlin [32], Hansen and Rand [33] treats the auto retrogressive coefficients and regression coefficients slopes as constant. ...
... Most works in this field test vector auto-regression coefficient (VAR) panel data models following the methods proposed by Holtz-Eakin et al. (1985, 1988, Weinhold (1996) and ...
This paper analyses the Granger-causality relationship between the growth of the real GDP per capita and the public debt, here represented by the ratio of the current primary surplus/GDP and the ratio of the gross Government debt/GDP. Using OECD annual data for 20 countries between 1988 and 2001, we adapt the methodology recently applied by Erdil and Yetkiner (2008) and we conclude that there is clear Granger causality and that it is always bi-directional. In addition, our findings point to a heterogeneous behaviour across the different countries. These results have important policy implications since not only does public debt restrain economic growth, but also real GDP per capita growth influences the evolution of public debt.
... We will follow the pioneer concept of Granger causality (Granger, 1969) and the approaches developed to analyse the existence of causality relationships among variables in panels (by such authors as Holtz-Eakin et al., 1988;Weinhold, 1996;Nair-Reichert and Weinhold 2001;Kónya, 2006;Hurlin and Venet, 2008), considering the general linear panel Granger causality model: ...
The relationships between bank market concentration and bank efficiency are of particular relevance in the European Union (EU), but they remain controversial. Using a panel Granger causality approach, this paper contributes to the literature, testing not only the causality running from bank market concentration to bank efficiency, but also the reverse causality running from efficiency to concentration. The results obtained confirm the relative complexity of these causality relationships, although they generally point to a negative causation running both from concentration to efficiency and from efficiency to concentration. These findings are in line with the Structure Conduct Performance (SCP) paradigm and the suggestions that the increase of the banks’ market power will contribute to inefficiency, since these banks will face less competition to obtain more output results with less input costs. Our results suggest that within this panel of all 27 EU countries over a relatively long time period, from 1996 to the onset of the 2008 financial crisis, the more cost-efficient commercial and savings banks operated in less concentrated markets.
... Most works in this field test vector auto-regression coefficient (VAR) panel data models following the methods proposed by Holtz-Eakin et al. (1985, 1988, Weinhold (1996) and ...
This paper analyses the Granger-causality relationship between the growth of the real GDP per capita and the public debt, here represented by the ratio of the current primary surplus/GDP and the ratio of the gross Government debt/GDP. Using OECD annual data for 20 countries between 1988 and 2001, we adapt the methodology recently applied by Erdil and Yetkiner (2008) and we conclude that there is clear Granger causality and that it is always bi-directional. In addition, our findings point to a heterogeneous behaviour across the different countries. These results have important policy implications since not only does public debt restrain economic growth, but also real GDP per capita growth influences the evolution of public debt. Key words: panel data; public debt and economic growth
... High income countries(28) ...
In this paper we investigate the causal relationship between financial development and economic growth. We use an innovative econometric method which is based on a panel test of the Granger non causality hypothesis. We implement various tests with a sample of 63 industrial and developing countries over the 1960-1995 and 1960-2000 periods. We use three standard indicators of financial development. The results provide support for a robust causality relationship from economic growth to the financial development. On the contrary, the non causality hypothesis from financial development indicators to economic growth can not be rejected in most of the cases. However, these results only imply that, if such a relationship exists, it can not be easily identified in a simply bi-variate Granger causality test.
This study explored the causal association between Information and Communications Technology (ICT) infrastructure development , measured by an ICT index, and the economic growth of 42 African economies between 2000 and 2019. Whether ICT development has contributed to real per capita GDP growth or ICT infrastructure expansion has just been a consequence of growth in real per capita GDP has been overlooked in prior studies concerning Africa. The econometric techniques used to analyse the data included robust second-generation tools to investigate cross-sectional dependence, slope homogeneity, and panel causality. The findings detected significant independence between the variables across countries, the slope was heterogeneous, and there was a long-run association between all of the economies in the sample. Dumitrescu and Hurlin's panel causality analysis indicated a uni-directional causal association between per capita income and the ICT index. The results also demonstrated that capital and employment were the leading causes of per capita GDP growth. The findings suggested that accelerating economic growth in developing economies was essential to promoting ICT investment. ARTICLE HISTORY
Few, if any, studies have examined the dual relationships between integration of developing countries into world technology markets and total factor productivity. This study uses a panel of 18 emerging markets and another of 20 developing economies to examine the bidirectional causal relationships between the growth rates of low, medium, and high technology exports, which represent integration into the world technology markets, and total factor productivity growth rates or productivity. The study uses data from 1990 to 2019 and the Granger causality panel data approach. Our empirical results confirm the following: First, there is a dual relationship between integration into the three world technology markets and total factor productivity growth rates at the level of both panels. At the country level, integration into the three world technology markets is the cause of productivity growth in emerging markets, but not in developing economies. Moreover, integration into higher technology markets leads to productivity gains in more emerging markets than integration into lower technology markets. On the other hand, total factor productivity growth rates does not cause the integration of emerging and developing countries into the three world technology markets. This study has two important policy implications. First, policy makers in developing countries need to develop the external technology sectors as key instruments to increase domestic productivity. Second, the accumulation of knowledge and know-how, without using them to integrate into world markets, would have little or no impact on productivity or sustaining economic growth in developing countries.
Une des originalités de notre approche cliométrique de l’histoire globale, a été de proposer une grille de lecture quantitative, multidisciplinaire et systémique. Notre grille est composée de cinq domaines interdépendants, désignés par l’acronyme ACIDE : Anthropologie, Cognition, Institutions, Démographie et Economie. Cette grille ACIDE est destinée à l’analyse multidisciplinaire des phénomènes économiques et en histoire globale. Dans le domaine institutionnel, nous avons défini deux indices originaux : l’un mesurant l’inclusivité économique et l’autre la non-inclusivité économique. Cela nous a permis de montrer de manière inattendue que la Chine a été plus inclusive économiquement que la moyenne de l’OCDE et la moyenne mondiale de 1995 à 2010. Même si l’inclusivité économique chinoise a chutée depuis lors, elle reste relativement proche de la moyenne mondiale. Une autre originalité de notre approche institutionnelle consiste à caractériser les institutions via, d’une part, des indicateurs « externes » liés à la souveraineté et la puissance, et d’autre part, des indicateurs « internes » de gouvernance et de respect des droits de l’individu. Nous définissons ainsi une matrice 2x2 de mesure des institutions regroupées en quatre catégories : les institutions politiques internes, les institutions politiques externes, les institutions économiques internes et les institutions économiques externes. L’analyse des dynamiques transitionnelles ACIDE, nécessite également de développer un temps transitionnel qui évolue au rythme de la transition étudiée et non du temps calendaire. Ce temps transitionnel permet de faire une analyse comparative de trajectoires de pays dont les transitions se déroulent à des époques, à des rythmes et à des intensités hétérogènes. Dans un domaine ACIDE donné, il est également nécessaire de détecter les phases historiques des trajectoires nationales et d’utiliser ces phases pour faire une analyse séquentielle des trajectoires. Notre apport original a été d’adapter l’analyse séquentielle, usuellement biographique, à l’analyse macroéconomique globale. Une des limites des analyses longitudinales est la disponibilité de données quantitatives riches sur le temps long. Notre solution originale à cette limite a été d’inférer les trajectoires moyennes sur le temps long, via une coupe transversale à une date donnée. Pour cela nous avons proposé une méthode permettant de définir et tester statistiquement un temps transitionnel pour le domaine étudié et qui permet alors de pallier l’absence du temps calendaire dans la coupe transversale. Avec le temps transitionnel ainsi créé, il est possible de caractériser une trajectoire transitionnelle moyenne en l’absence de temps calendaire. Il est intéressant d’utiliser l’aspect universel et propagatoire de la transition dans un domaine disciplinaire ACIDE donné, pour effectuer des prévisions actuarielles de long terme pour les pays à transition plus tardive, en prenant en compte, non seulement leurs trajectoires propres, mais également l’historique des trajectoires des pays à transition précoce. Notre solution originale à ce besoin passe par la création de variables exogènes dites « variables cliométriques » qui permettent de faire le lien statistique entre les deux types de pays. Nous appliquons notre technique dans le domaine démographique pour les prévisions actuarielles à long terme des taux de mortalité par âge et sexe. Les modèles généralisés et optimisés que nous avons introduits ont permis d’améliorer les modèles actuariels classiques à l’horizon de 50 ans. Pour le pays émergent étudié (Inde), la prise en compte des « variables cliométriques » a abouti à des prévisions beaucoup plus précises à l’horizon de 40 ans. Enfin, nous appliquons la grille cliométrique multidisciplinaire ACIDE à l’étude des déterminants multidisciplinaires de la mortalité dans l’infanterie française, au cours de la première Guerre mondiale.
This study examines the validity of the export-led growth hypothesis in the context of developed countries’ agricultural sector, considering the existence of a common market among countries. Using the annual data of Organisation for Economic Co-operation and Development (OECD) countries from 1997 to 2016, we analyze the effect of agricultural exports on agricultural growth. For the comparison between developed countries with and without common markets, we create three subsamples (OECD, the European Union [EU] countries among OECD members, and the non-EU countries among OECD members). Estimation results show that agricultural exports (imports) have a positive (negative) effect on agricultural growth only in the case of the EU subsample. These results imply that access to foreign markets is important to validate the export-led growth hypothesis in the agricultural sector of developed countries.
The relationship between financial development and economic growth is among the essential issues
analyzed in economy literature. This study researches the relationship between financial
development and economic growth by using 1980-2014 period annual data for Australia, Denmark,
Chile, Iceland, Israel, Japan, Korea, Switzerland, Turkey and Mexico. As the indicator of financial
development, the proportion of domestic credits provided by financial sector to Gross National
Product was used as the indicator of economic growth the growth rate of Gross National Product was
used. Panel causality test developed by Dumitrescu – Hurlin (2012) was utilized in order to research
the relationship between variables. While the obtained findings indicate the existence of a causality
relationship from financial development to economic growth for the countries determined, a causality
relationship from economic growth to financial development could not be identified.
The impact of unionization on wage inequality has been examined by a vast literature. Focusing mostly on the US and the UK in time series analyses or on OECD countries in panel data analyses, a bulk of these studies have found a negative impact of deunionization (i.e. decline in the union density rate) on distribution of wages. By utilizing two inequality data sets both provided by the University of Texas Inequality Project and the union density data set provided by OECD this paper contributes to the literature, analyzing the causality relationship between deunionization and pay inequality for 24 OECD countries for the 1963–2000 period within a panel Granger structure. Our findings show not only that there is causality from union density to income inequality but also, perhaps more importantly, point out that there is causality running from income inequality to union density for various set of countries and time periods.
Dynamique périurbaine et dépenses publiques locales : une analyse en termes de causalité
par Marie-Estelle Binet
Cet article propose une méthodologie économétrique originale afin de tester l'existence de compétition fiscale entre collectivités locales, via la croissance des dépenses publiques locales. Une analyse de la causalité au sens de Granger entre le niveau de population résidentielle et le montant de bien public offert dans une localité est ainsi menée sur un échantillon de 25 communes françaises sur la période 1960-1983. On montre que, pour 8 communes étudiées, le niveau de bien public local précède et influence positivement les choix d'implantation. La croissance résidentielle de ces communes est donc expliquée, en partie, par une stratégie de compétition fiscale. Pour 7 autres communes, la causalité inverse est observée. Elles sont donc caractérisées par un comportement adaptatif.
Existing studies examining the Granger causality relationship between energy consumption and GDP use a panel of countries but implicitly assume that the panels are homogeneous. This paper extends the Granger causality relationship between energy consumption and GDP by taking into account panel heterogeneity. For this purpose, we use a large panel of 79 countries for the period 1980–2007. Specifically, we examine four different causal relationships: homogeneous non-causality, homogeneous causality, heterogeneous non-causality, and heterogeneous causality. The results show that roughly seven-tenths of the countries exhibit bi-directional Granger causality, two-tenths exhibit no Granger causality, and one-tenths exhibit uni-directional Granger causality.
Does the institutional environment affect the causal relationship between banking development and economic growth? In the theoretical section of this paper, we develop an endogenous growth model where the institutional environment is captured through two indicators: judicial system efficiency and easiness of informal trade. We show that an improvement in the institutional environment has two effects. First, it intensifies the causality direction from banking to economic growth through a reduction in defaulting loans. Second, it reduces the interest rate spread. In the empirical section of the paper, we find bidirectional causality when analyzing 22 Middle Eastern and North African countries over the period 1984–2004. The first causality, which runs from banking development to economic growth, is more intense in countries with more developed institutional environment. The second causality runs from economic growth to banking and indicates that a more developed economy has a more developed banking system.
This study investigates the Granger-causality relationship between real per capita GDP and real per capita health care expenditure by employing a large macro panel data set with a VAR representation. The findings verify that the dominant type of Granger-causality is bidirectional. In instances that we found one-way causality, the pattern is not homogenous: Our analyses show that one-way causality generally runs from income to health in low- and middle-income countries whereas the reverse holds for high-income countries. Accordingly, care must be paid in defining the dependent and independent variables when specifying the determinants of health care expenditure.
In this paper, we propose a simple Granger causality procedure based on Meta analysis in heterogeneous mixed panels. Firstly, we examine the finite sample properties of the causality test through Monte Carlo experiments for panels characterized by both cross-section independency and cross-section dependency. Then, we apply the procedure for investigating the export led growth hypothesis in a panel data of twenty OECD countries.
In economic development, aggregate economic growth is accompanied by structural change among the three main sectors of an
economy. Nevertheless, the question whether economic growth causes structural change, or changes in the economic structure
cause aggregate growth is still unanswered. To shed more light on this question, this article examines a Granger causality
test in a panel environment to determine the relationship of economic growth and structural change, measured either in terms
of employment shares or real value added shares. Estimation and analysis of the annual data of seven OECD countries, covering
the period from 1960–2004, show that although the causality appears to be heterogeneous among these countries, some general
conclusions can be drawn. Aggregate economic growth decelerates structural change in the very short run but accelerates it
with some lag in time. The aggregate effect depends on whether structural change is measured in terms of employment or in
terms of real value added. Conversely, structural change supports aggregate economic growth, irrespective of which measure
of structural change chosen.
KeywordsStructural change–Economic growth–Tertiarization–Panel Granger causality test
Testing Granger causality in heterogeneous panel data models
This paper proposes a simple test of Granger [1969] non causality hypothesis in heterogeneous panel data models with fixed coefficients. It proposes a statistic of test based on averaging standard individual Wald statistics of Granger non causality tests. First, this statistic is shown to converge sequentially to a standard normal distribution. Second, for a fixed T sample the semi-asymptotic distribution of the average statistic is characterized and a standardized statistic based on general approximations of the moments of individual Wald statistics is proposed. Monte Carlo experiments show that both statistics provide good small sample properties.
Many are the factors indicated by the pertinent literature concerning the causes of deforestation in Brazilian Legal Amazon. From endogenous aspects as the edafo-climatic conditions to aspects related to anthropic action, like the population movements, urban growth, and especially, the independent or induced actions of the different public and private economic agents who have acted in the region, historically configuring the processes of occupation of the land and economic exploitation of the Amazonian region. The objective of this article is to perform a causality test, in the Granger sense, in the main variables suggested as important that explain the deforestation of the Legal Amazon, in the period from 1997 to 2006. The methodology to be used is based on dynamic models for the panel data, developed by Holtz-Eakin et al. (1988) and Arellano-Bond (1991) who developed a causality test based on the seminal article of Granger (1969). Among the main results found is the empirical evidence that there is a bidirectional causality between deforestation and the areas of permanent and temporary cultures, as well as the size of the cattle herd.
This paper proposes a very simple test of Granger (1969) non-causality for heterogeneous panel data models. Our test statistic is based on the individual Wald statistics of Granger non causality averaged across the groups. First, this statistic is shown to converge sequentially to a standard normal distribution. Second, for a fixed T sample the semi-asymptotic distribution of the average statistic is characterized. A standardized statistic based on an approximation of the moments of Wald statistics is proposed. Monte Carlo experiments show that our panel standardized statistics provide very good small sample properties.
The remarkable increase in FDI flows to developing countries over the last decade has focused attention on whether this source of financing enhances overall economic growth. We use a mixed fixed and random (MFR) panel data estimation method to allow for cross country heterogeneity in the causal relationship between FDI and growth and contrast our findings with those from traditional approaches. We find that the relationship between investment, both foreign and domestic, and economic growth in developing countries is highly heterogeneous and that estimation methods which assume homogeneity across countries can yield misleading results. Our results suggest there is some evidence that the efficacy of FDI in raising future growth rates, although heterogeneous across countries, is higher in more open economies. Copyright 2001 by Blackwell Publishing Ltd
The relationships between the variables measuring socio-economic status (SES) and those measuring various aspects of health and mortality were analyzed. Various causality definitions were used with statistical data, which was considered a vector of time series, for the analysis. The SES variables considered in the analysis included wealth, income, education, unsafe neighbourhoods and dwelling poor/fair. The use of association information between health and SES variables for forecasting values of health variable was also discussed.
ResearchGate has not been able to resolve any references for this publication.