Contenu du sommaire : Structures et propriétés de cinq modèles macro-économiques français

Structures et propriétés de cinq modèles macro­économiques français

• Avant-propos - p. 5
• Présentation générale - p. 1-14
  Structure and Properties of Five French Macroeconometric Models This paper compares the structure, main specifications and properties of large-scale macroeconomic models regularly used for macroeconomic forecasts and evaluations by five French institutions (Banque de France, Direction de la Prévision, Ecole Centrale, Insee and OFCE). The main differences between the results obtained with each model are due to the choice of modeling by field or variable. These particularities are described by breaking down the models into four main blocks covering all the relations concerning a particular field. The first block represents business behaviour: production functions, input demands and inventory build-up equations. In four of the five models, the functions used have complementary factors. The accelerator principle plays a central role in investment, but other variables such as profits, profitability and pressure on production capacities also frequently intervene. The employment dynamic results from adjustment to a target, which depends on the underlying productivity trends and production prospects. The mechanism for inventory build-up by businesses assumes stock adjustment to a chosen level, which depends on production. Buffer stock effects are also taken into account as is the effect of speculative behaviour on price trends. The second block concerns the price-wage loop in the broad sense: wage setting, price setting and unemployment modeling. As regards wages, virtually all the models opt for the hypothesis of a Phillips relation with long-term unit indexation to price increases. The prices are modeled on the basis of mark-up patterns. Unemployment is determined by comparing the labour force with employment. All of the models take into account the reaction of the labour supply to the labour market situation (activity rate fluctuation mechanism). The third block deals with the how households divide their income between consumption and savings and their choice of housing investment level. Most of the equations used in the models assume the long-run unit income elasticity of consumption. Differences are found mainly in the precise definition of the disposable income subject to the savings-consumption choice and the inclusion of inflation effects (real cash balances effect) and other variables characteristic of the general economic situation (especially unemployment). When household housing investment behaviour is endogenous, the investment is assumed to adjust the actual stock to the desired stock level, which depends on demography, income and sometimes the real interest rate. The fourth block presents the foreign trade equations. Exports and imports are modeled in very similar ways and the equations used ascribe a central role to price competitiveness and growth in national and foreign demand.
• La demande des entreprises sur le marché des biens et le marché du travail - p. 1-19
• La boucle prix-salaires et la détermination du chômage - p. 21-40
• La demande des ménages - p. 41-56
• Le commerce extérieur de produits manufacturés - p. 57-70
• Sensibilités des taux de change d'équilibre aux output gaps et aux cibles de la balance courante - Florence Thibault, Cécile Couharde, Didier Borowski p. 71-96
  Sensitivities of equilibrium exchange rates to output gaps and current-account balance targets. Methodology and estimates for the main industrialized countries by Didier Borowski, Cécile Couharde and Florence Thibault Following the research initiated by the IMF in the 1970s, John Williamson defined the concept of fundamental equilibrium exchange rate (FEER) as the real effective exchange rate that enables an economy to achieve two simultaneous goals in the medium term: internal equilibrium (i.e., a non-accelerating-inflation growth path) and external equilibrium (i.e., a "sustainable" current-account balance). The aim of this article is to estimate FEERs for the main industrialized countries (G7 + Belgium, the Netherlands, and Spain) over the period 1984-95. To obtain these estimates, we constructed a static model based on the long-term specifications of the National Institute Global Economic Model (NIGEM). Our approach is partly inspired by the work of the National Institute of Economic and Social Research (NIESR) but departs from it to the extent that we assume internal equilibrium to be exogenous. The main determinants of FEERs are the current-account balance targets and potential GDP growth rates — which define the external and internal equilibriums respectively. Given their high normative content, the values we assign to both determinants are essentially intended as notional examples. Our method for resolving the long-term model offers two major advantages: (1) We directly obtain bilateral equilibrium parities from which we calculate the real effective exchange rates. (2) We demonstrate the sensitivity of exchange rates to deviations from current-account balance targets and to output gaps. This approach thus allows a more accurate measurement of currency over- or undervaluation .