Macromodels of the Romanian transition economy, Second edition

The book develops the main lines of thought contained in “Macromodels of the Romanian transition economy”, edited in 1996 by the “Expert Publishing House”. The 1997 version of the macromodel (Dobrescu 1997 b) has included some changes, the following being the most important: - the introduction of the special block for demographic variables (population, labour force, retired people); - the connection of the annual indicators with a monthly block dedicated to the evolution of export and exchange rate; - the aggregation of the previous five sectors in the following three: a) industry, construction and agriculture; b) transport, communication, trade, banking and other services; c) public services; - the re-estimation of the econometric functions on the basis of updated statistical series (including the provisional data for 1996). The 1998 version of the macromodel, presented in this book, contains new improvements: - a more relevant determination of the expected disposable income of households, firms, and general consolidated budget; - the delimitation of the main consequences of the budget deficits; - a more detailed elaboration of the possible scenarios of the future evolution of the Romanian economy. The present book tries to define the features of the weakly structured economy from an institutional perspective. The institutional framework is studied from three points of view: a) the main components (property rights, rules of human interaction, the amplitude and ways of the discretionary intervention of public authorities in the economic life); b) the degree of specification of these components (clearly and uncontradictorily defined, ambiguously defined); c) social validation (formal or informal). Starting from the possible combinations of these elements, two types of economic systems can be distinguished: the first implies a high expected stability and is denoted as a structured economy and the second is characterised by a relatively low expected stability, being considered a weakly structured economy. The transition economy, at least in the case of Romania, is defined as weakly structured: the property rights are not yet clearly delimited; the economic life is marked by the mixture of old and new rules and organisations involved in human interaction; discretionary intervention of the public authorities is very large and submitted to random political interests; the formal institutions are incomplete and soft, but the informal ones have an important role in economy and society. On this theoretical basis, the main macroeconomic implications are analysed: a) chronically inefficient utilisation of the production factors; b) persistence of inter-enterprise arrears and of disturbing form of “dollarization”; c) large share of non-accounted eco-nomy; d) monetary distortion and asymmetry of liquidities. The weakly structured economy is characterised by congenital instability and, therefore, the modelling problems are especially complicated. The notion “econometric model” is used in the following meaning: as a set of interdependent equations (from which at least one is econometric) approximating a particular given class of statistical data in accordance with the modeller’s image about functional relations among respective series. If the model reflects a “given class of statistical data”, it is evident that it can be used only for the analysis of this information; forecasts are acceptable exclusively in the proximity of the respective time interval. On the other hand, the “image” represents a mixture of theoretical assumptions adopted (explicitly or implicitly) by the modeller, and also of his beliefs, intuitions, attitudes and desires concerning the studied process. Consequently, for every economic system a large variety of models are possible depending on the conceptual premises of their creators. Maybe, this relativism is intellectually uncomfortable, but it is inherently implied in econometric modelling, especially when a weakly structured economy is approached. The most difficult problem is the stationarity of statistical data. In order to obtain an overview about this question, 76 annual and 14 monthly series have been exposed to Augmented Dickey-Fuller Test. The basic series and their natural logarithms are stationary only in 34% of the cases for annual data; the monthly data are better situated (68%), but they are relevant for very few correlations. The general opinion about frequent stationarity of the first and second differences are confirmed. Instead, the indices and the corresponding rates are less stationary. The best performance is registered by the first difference of indices and their variation. Under these conditions, two modelling approaches are possible: a) to use, partially at least, the basic unstationary series, the stability of macromodel usually being higher than the stability of separate functions as a result of the interactions among them and the accounting identities (a similar solution has been adopted for the 1996 version of the macromodel); b) to use only stationary series, that is preponderantly derived indicators with supplementary problems (in forecasts) generated by their translation in basic ones (the 1997 and 1998 versions are built on this principle). The appendices of the book contain a set of the most relevant macroeconomic indicators of Romania for 1980-1996 (annual data) and January 1991 - December 1996 (monthly data), the detailed presentation of the econometric functions, the main scenarios of the Romanian eco-nomy for 1998-2000, a selected bibliography and thematic index.