L'énergie et l'argent d'un nouveau cadre de modélisation macroéconomique

Depuis la stagflation observée consécutivement à la forte hausse du prix du pétrole en 1973 et 1979, les chocs pétroliers sont considérés comme l’une des sources de fluctuations potentiellement les plus importantes aux États-Unis comme dans de nombreux pays industrialisés. De nombreux articles ont étudié le rôle des chocs pétroliers dans la fluctuation des principales variables macroéconomiques à savoir, la croissance, le chômage, l’inflation et les salaires. Cependant, ces travaux n’ont pas encore permis d’aboutir à un consensus. Le débat s’est même intensifié au cours de cette dernière décennie, en raison d’une absence de réaction forte de l’économie réelle pendant la période d’augmentation du prix du pétrole entre 2002 et 2007. En effet, la récession qu’aurait dû engendrer une telle hausse des prix ne fut observée qu’au moment de la crise des subprimes en 2008. Plusieurs hypothèses furent avancées pour expliquer la différence entre les crises des années 1970 et 2000. Blanchard & Gali (2009) et Blanchard & Riggi (2013) évoquent, par exemple, la réduction de la quantité de pétrole utilisée dans la production, la plus grande flexibilité des salaires réels et une meilleure crédibilité de la politique monétaire. Hamilton (2009) et Kilian (2008) suggèrent quant à eux de l’expliquer par l’origine différente des deux chocs pétroliers : un choc d’offre pendant les années 70 et un choc de demande pendant les années 2000. L’objectif original de la thèse était de réexaminer l’impact des chocs pétroliers sur l’économie réelle par le canal de la dette. [...]Le développement de ces travaux entamés dans la thèse pourra aboutir à un cadre alternatif de modélisation décisif pour l’intelligence de la macroéconomie. Il devrait permettre une meilleure compréhension de l’évaluation des relations réciproques entre la sphère financière, la réalité des cycles macroéconomiques réels, l’énergie et le climat dans ce qui est sans aucun doute l’enjeu de notre génération : la transition écologique.Ever since the stagflation that followed the oil price run-ups of 1973 and 1979, oil price shocks have been considered one of the most influential sources of economic fluctuation in the United States and other developed countries. A large body of literature has analyzed oil price shocks as sources of variation for leading macroeconomic variables such as GDP growth, unemployment rate, inflation, and wages. However, scholars have yet to reach a consensus as to the true impact of oil shocks on the macroeconomic environment. Furthermore, the last decade has seen the debate intensify as the results of the relatively (in comparison with the 1970s) muted reaction of the real economy during the 2002-6 oil price run-up. Indeed, the recessionary effect was only observed during the subprime mortgage crisis of 2008-9. Numerous hypotheses have been put forward to explain the difference in impact during the 1970s versus the 2000s. For instance, Blanchard & Gali (2009) and Blanchard & Riggi (2013) evoked the reduction of the quantity of oil used of a unit of production, more flexible real wages, and a better credibility of the monetary policy. Hamilton (2009) and Kilian (2008) pinpointed a difference in the nature of the shock: whereas the oil shocks of the 1970s were driven by supply, that of the 2000s was led by demand. The original aim of this thesis was to reevaluate the impact of the oil shock in the 2000s through the debt channel. First, based on the work of Banchard & Gali, we proposed a new dynamic stochastic general equilibrium model (DSGE), which includes oil as an input of production as well as a consumption good. By relaxing some of the hypotheses of Blanchard & Gali, especially the decoupling of the output elasticity of oil with the cost-share in the production, our work demonstrated that oil is still a fundamental variable of the GDP in the United States. Furthermore, we found that energy efficiency is a key factor that explains the muted macroeconomic impact of an increase in oil prices. A third line of inquiry that may explain the difference between the shocks of the 1970s and the 2000s considers the extra costs implied by a higher price of oil that were absorbed by private debt (which was itself exacerbated by low interest rates set by the Federal Reserve in the 2000s). However, we found that DSGE modeling is unable to replicate the macroeconomic environment that led to the subprime mortgage crisis. In light of these considerations, I reoriented my thesis along the lines of a new angle of research that seeks to represent economic mechanisms differently. Under this new frame-work, private debt is at the core of macroeconomic analysis. It provides an alternative view of the financial crisis that occurred in the 2000s.[...]The conclusions of this thesis demonstrate great potential for providing foundations for new perspectives in macroeconomic modeling. The papers included in the thesis allow, in particular, for a better understanding of situations that most macroeconomic models are not able to cope with, including the over indebtedness crisis. As a result, the framework introduced here may provide an alternative and improved perspective for public policy. Further development of the research presented in this thesis may lead to the improvement of other frameworks in the field of macroeconomics. This would allow for a better understanding of complex interactions between the financial sphere, real business cycles, energy, and climate in what is certainly the biggest challenge of our generation : the ecological shift