Volume 214, Issue 6

This paper tries to give a formal proof of the Coase Theorem. In a first step it is shown in a general bargaining model without bargaining costs, that the efficiency of a solution depends on the strategies used by the parties (stationary or non-stationary). By transforming the game into a modified Rubinstein-Game, a unique and efficient solution is obtained. In this, however, there is still a potential for strategic threats (extortion). If the parties use these threats, multiple equilibria will result, some of which are inefficient. Therefore, even without bargaining costs, the Coase Theorem is not generally valid.

A large and expanding part of recent economic data sets are produced with the help of analytical models and therefore have the character of hypotheses. Constructing analytical models on the basis of such model-generated data leads to a “modeling on the basis of the results of modeling”. In this aricle the authors show that this procedure has consequences that seriously endanger the quality of those studies. To avoid further unreflected progagation of this type of empirical research they propose a “Charter for Compilation and Correct Handling of Economic Data”.

This paper is intended to explore a new way of analyzing macroeconomic datasets. For a dataset of twenty-one West German variables a common-trends-common-cycle representation is aimed at which allows for a multivariate Beveridge-Nelson decomposition of the data. However, standard F-tests reveal that the number of linearly independent cointegrating vectors plus the number of linearly independent serial correlation cofeature vectors do not add up to the dimension of the system. A slightly weaker notion of cofeature is used to derive an approximate trend-cycle decomposition. For most of the variables, this approximation seems to work quite well. Analysis of the cyclical components indicates that interest rates are procyclical with a lead and anticyclical with a lag. The real wage is at most very weakly procyclical and definitely not anticyclical, while the price level is procyclical. Zusammenfassung In diesem Aufsatz wird eine neue Methode zur Bestimmung der Beveridge-Nelson-Zerlegung untersucht. In einem Modell gemeinsamer Trends und gemeinsamer Zyklen werden kointegrierende Vektoren und sogenannte Cofeature-Vektoren für einundzwanzig westdeutsche Makrovariablen bestimmt. Mit einer leichten Abschwächung des Cofeature-Konzeptes kann eine approximative Trend-Zyklus-Zerlegung bestimmt werden. Die Approximation scheint für die meisten Variablen von recht guter Qualität zu sein. Ausgewählte Ergebnisse beinhalten, daß der Reallohn im wesentlichen azyklisch ist, das Preisniveau prozyklisch und Zinssätze prozyklisch mit einem Lead, jedoch antizyklisch mit einem Lag sind.

The paper uses the so-called Minimal Flow Analysis (MFA) Method as an efficient tool of production-structure analysis. It applies a graph-theortical approach which remedies to some extent existing deficits of other Qualitative Analysis Methods by increasingly taking into account quantitative information. Exemplaric use of MFA on an interregional input-output table including 7 EC-countries reveals some perhaps unexpected, but very plausible parallels between some of the countries. This also suggests some hypotheses concerning structural development. Zusammenfassung Das Papier zeigt unter Anwendung der sog. Minimal Flow Analyse (MFA)-Methode die Analyse der Produktionsstruktur. Der verwendete graphentheoretische Ansatz verbessert einige Schwächen anderer Qualitativer Analysemethoden, indem er quantitative Informationen auf verschiedenen Vorstufen der Produktion verstärkt in Betracht zieht. Die exemplarische Anwendung der Methode auf eine interregionale Tabelle von 7 europäischen Ländern enthüllt einige unerwartete aber dennoch äußerst plausible Parallelen in der Produktionsstruktur einiger Länder. Diese Ergebnisse könnten auch als Ansatzpunkt für eine Hypothesenbildung der Strukturentwicklung genommen werden.