This thesis is concerned with the development and use of a macroeconometric model for the Algerian economy between 1963 and 1984. The model was built because of a systematic lack of applied econometric studies pertaining to Algeria at both the macroeconomic and microeconomic level. It is hoped that the model will fill a gap in this area and will contribute to the much neglected field of applied econometric research with regard to Algeria. This lack of applied econometric studies for Algeria meant that the modelling exercise described here has had to rely on an extensive specification search based on evidence relating to Algeria's economic structure and policy, economic theory, and the experience of Less Developed Countries in the area of macroeconomic model-building. The lack of data was a major constraint in this area and part of this study consisted of collecting and compiling a large database. After the country's independence in 1962, Algerian macroeconomic policy aimed to create a strong industrial system and to satisfy the population's basic needs. It relied on heavy industrialisation to modernise the economy, oil revenues to finance development, and central planning as the major tool of macroeconomic regulation. The accumulation rate was high and the growth record was generally good. However high unemployment and inflation, considerable disequilibrium, low productivity, a vulnerable balance of payments and unsustainable external debt are the major macroeconomic problems that policy-makers have had to face. The model's equations were first estimated using the OLS method and were subjected to stringent statistical tests. The degree of test significance and parameter correspondence to a priori views on the economy was good. when the model was constructed, it was estimated using a 2SLS principal component method. The OLD results were found to be reasonably feasible. The equations were collected into a system of 63 equations and solved using dynamic simulation technique. The model was solved successfully and its tracking of historical data was reasonably good. Further tests were carried out to study its dynamic features. Having constructed the model, it was then used extensively to perform simulation analysis. The experiments ranged from those concerning the goverment's current expenditure to its monetary policy. In all, nine simulation exercises were carried out. These were revealing on the workings of the Algerian economy. The model was further used in scenario analysis. First the model was used to develop an ex ante forecast employing a linear trend model for the exogenous variables. The forecast database was used to generate multipliers. The policy analysis was constructed to coincide with the implementation of the Second Five Year Plan (1985-1989). The feasibility of the plan was examined by varying the price of oil according to three hypotheses. The aim of this test was to develop a realistic framework for applied macroeconomic analysis.

This thesis is concerned with the development and use of a macroeconometric model for the Algerian economy between 1963 and 1984. The model was built because of a systematic lack of applied econometric studies pertaining to Algeria at both the macroeconomic and microeconomic level. It is hoped that the model will fill a gap in this area and will contribute to the much neglected field of applied econometric research with regard to Algeria. This lack of applied econometric studies for Algeria meant that the modelling exercise described here has had to rely on an extensive specification search based on evidence relating to Algeria's economic structure and policy, economic theory, and the experience of Less Developed Countries in the area of macroeconomic model-building. The lack of data was a major constraint in this area and part of this study consisted of collecting and compiling a large database. After the country's independence in 1962, Algerian macroeconomic policy aimed to create a strong industrial system and to satisfy the population's basic needs. It relied on heavy industrialisation to modernise the economy, oil revenues to finance development, and central planning as the major tool of macroeconomic regulation. The accumulation rate was high and the growth record was generally good. However high unemployment and inflation, considerable disequilibrium, low productivity, a vulnerable balance of payments and unsustainable external debt are the major macroeconomic problems that policy-makers have had to face. The model's equations were first estimated using the OLS method and were subjected to stringent statistical tests. The degree of test significance and parameter correspondence to a priori views on the economy was good. when the model was constructed, it was estimated using a 2SLS principal component method. The OLD results were found to be reasonably feasible. The equations were collected into a system of 63 equations and solved using dynamic simulation technique. The model was solved successfully and its tracking of historical data was reasonably good. Further tests were carried out to study its dynamic features. Having constructed the model, it was then used extensively to perform simulation analysis. The experiments ranged from those concerning the goverment's current expenditure to its monetary policy. In all, nine simulation exercises were carried out. These were revealing on the workings of the Algerian economy. The model was further used in scenario analysis. First the model was used to develop an ex ante forecast employing a linear trend model for the exogenous variables. The forecast database was used to generate multipliers. The policy analysis was constructed to coincide with the implementation of the Second Five Year Plan (1985-1989). The feasibility of the plan was examined by varying the price of oil according to three hypotheses. The aim of this test was to develop a realistic framework for applied macroeconomic analysis. / Algerian Ministry of Higher Education

Cette thèse prend pour objet les transformations de l’analyse macroéconomique aux États-Unis durant les années 1970 tout en questionnant la manière d’étudier et d’analyser ces transformations. Du point de vue de l’histoire des faits, la période semble marquer une rupture par rapport à la relative stabilité économique de l’après-guerre. Cette période d’instabilité économique, qu’on nomme stagflation, fait écho à l’instabilité de la théorie macroéconomique aux États-Unis. Le consensus de l’époque, considéré comme « keynésien », se retrouve attaqué par les économistes dits « monétaristes » et « nouveaux classiques ». Le dernier des groupes cités est celui des « révolutionnaires », celui dont on considère qu’il a changé radicalement la discipline. Le but de ma thèse est d’étudier l’influence des nouveaux classiques sur la macroéconomie dans les années 1970 en mobilisant un appareil historiographique qui met au cœur de l’étude le rôle joué par la stagflation, et de confronter les résultats de cette étude avec l’histoire « conventionnelle » de la macroéconomie. La thèse s’articule autour de quatre articles indépendants les uns des autres. Le premier chapitre propose une comparaison entre les méthodologies de Lucas et Sargent, et montre que le second tente de donner un caractère plus réaliste aux modèles de la Nouvelle Économie Classique, en utilisant les anticipations rationnelles pour décrire différents phénomènes économiques. Le second chapitre prend pour objet la confrontation entre Lucas et Sargent d’un côté, et les défenseurs des modèles macroéconométriques structurels de l’autre. Le chapitre 3 étudie l’évolution des travaux de Robert Gordon sur l’inflation dans les années 1970 et documente la manière dont celui-ci adopte petit à petit l’hypothèse de taux de chômage naturel. Le chapitre 4 enfin s’intéresse aux débats empiriques au début des années 1980, autour de la crise de Lucas. / This thesis focuses on the transformations of macroeconomics in the United States during the 1970s, while questioning the way to study and to analyze these transformations. From the point of view of economic history, the period seems to mark a break with the relative stability post World War II years. This period of economic stability, that one calls “stagflation”, echoes the instability of U.S. macroeconomic theory. The consensus of the time, regarded as “Keynesian”, is attacked by economists labeled as “Monetarist” and “New Classical”. The last group is the one of “revolutionaries”, regarding as having radically transformed the discipline, as the Copernican revolution overthrown the geocentric representation of the universe. My goal in the thesis is to study the influence of New Classical economists on macroeconomics in the 1970s, by appealing to an historiographical framework which outs at the heart the role played by stagflation, and by confronting the results of this work to the standard narrative. This thesis is built around four articles, independent from one another. The first chapter proposes a comparison between the methodologies of Lucas and Sargent, and shows how the latter intend to give a more realistic character to the new classical economy models, by using rational expectations to describe different economic phenomena. The second chapter takes interest in the confrontation between Lucas and Sargent on one side, and the defenders of structural econometric models on the other. The third chapter studies the evolution in the works of Robert Gordon on inflation in the 1970s, and documents the way he gradually adopts the natural rate of unemployment hypothesis. Finally, the chapter four is interested in the empirical debates in the early 1980s, about the Lucas critique.

Le Luxembourg est le 3ème exportateur mondial de services financiers. Il figure parmi les pays qui accueillent le plus d’investissements directs en provenance de l’étranger, ce qui indique l’intensité de ses liens avec l’économie mondiale. Le but de ce travail est d’analyser l’influence éventuelle d’une économie mondiale caractérisée par l’interdépendance des marchés réels et financiers sur l’économie Luxembourgeoise. Chapitre 1 présente une analyse des interactions de l’économie Luxembourgeoise avec le monde extérieur. Le chapitre suivant est consacré à la revue de la littérature portant sur la modélisation de l’intermédiation financière au niveau macroéconomique, couvrant plusieurs types d’approches de modélisations. Enfin, le troisième chapitre comporte un modèle macroéconométrique multi-pays construit et analysé afin de simuler les scénarios plausibles. Le modèle y est présenté avec ses fondements théoriques, les résultats des simulations et une comparaison avec d’autres modèles. La nouveauté du modèle réside dans sa prise en compte du commerce international désagrégé en services financiers et autres, et des investissements internationaux en portefeuille avec leurs flux de titres et de capitaux, ainsi que de leur impact sur la croissance économique. Les résultats des simulations montrent que ce cadre d’analyse donne parfois des résultats différents par rapport aux modèles standards. Nombre de scénarios qui ne peuvent être simulés par d’autres modèles, tels que la baisse des flux internationaux d’investissements de portefeuille, sont également analysés et confirment la forte vulnérabilité du Luxembourg aux chocs externes qui ont lieu sur les marchés financiers. / Luxembourg is world’s third financial services exporter and one of world’s top recipients of foreign direct investment in value as well as per capita terms, which highlight its strong linkages with world economy. The objective of this dissertation is to analyze possible outcomes for the very small and very open economy of Luxembourg in a world environment, where real and financial markets affect each other. To better understand the characteristics of the economy and economic mechanisms behind them, a thorough analysis with emphasis put on the interactions with the outside world based on available data and relevant literature is presented (Chapter 1). Subsequently a survey of literature devoted to modeling financial intermediation at macroeconomic level across various types of modeling approaches is offered (Chapter 2). Finally, a multi-country macroeconometric model built to simulate possible scenarios is presented and analyzed (Chapter 3) with its theoretical background, simulations’ results and comparison with other models. The model is novel in that it accounts for international trade disaggregated into financial services and the rest, and international portfolio investment in securities and equity flows, that have a significant impact on the country’s economic growth. Simulations’ results show that such a framework generates sometimes markedly different results than more standard models. A number of scenarios which cannot be simulated in other models, such as American stock market fall or a decrease in international portfolio flows, are also analyzed and confirm the high vulnerability of Luxembourg economy to external shocks originating in financial markets.