Petre and Corina Saman
Published in Romanian Journal of Economic Forecasting,
- Requires a PDF viewer such as
or Adobe Acrobat Reader
this article we present only the economic forecast of the variable
of interest; for a description of the model see the previous number
of the Journal .
The macromodel estimates the short and medium-term economic implications for internal policies and of changes in the international context.
This new version of the Romanian macromodel incorporates the experience accumulated through the utilisation of its previous forms - either experimental (tested during 1991-1995) or operational (developed during 1996-2003). At the same time, it introduces some methodological and information improvements.
The most significant of them is the structural decomposition of the economy, associated with input-output techniques. Output and absorption are divided into: a) agriculture, sylviculture, forestry, hunting, and fishing; b) mining and energy; c) manufacturing industry; d) constructions; e) transport, post and communications; f) trade and services. These can be easily translated into classical three-sector classification: primary (a-b), secondary (c-d), and tertiary (e-f).
Due to the relatively advanced stage of the transitional processes in Romania, the behavioural functions were accommodated - as much as possible - to the standard relationships. Unlike the versions that used the statistical series beginning with 1980, the present one is based exclusively on information concerning the period 1989-2004. Therefore, we have considered more adequate to name this variant the macromodel of the Romanian market (not transition, as before) economy.
Since the input-output tables are defined yearly, the model contains only annual indicators. They are expressed in denominated local currency (RON). When there were several statistical sources for the same indicator, we preferred the data extracted or derived from national accounts.
The statistical series are relatively short and often fractured (because of the transformation processes of transition). It is known that ADF test of stationarity does not offer sure results in the case of limited number of observations. Nevertheless, the series satisfying it were used, as a rule. The Granger causality test was computed for one, two, and three lags. The simplest methods of estimation were also preferred. The structural breaks in the evolution of some indicators were attenuated by the inclusion of dummies. Obviously, all these circumstances weaken the stability of the econometric coefficients that must be continuously updated. The main relationships are grouped in seven sections: input-output block; labour market, production function; domestic absorption, foreign trade, prices and exchange rate, and interest rate.
input-output analysis, econometric relationships, simulations
JEL Classification: C5, E2, E6, H6