To promote transparency and provide information, the Federal Planning Bureau regularly publishes the methods and results of its works. The publications are organised in different series, such as Outlooks, Working Papers and Planning Papers. Some reports can be consulted here, along with the Short Term Update newsletters that were published until 2015. You can search our publications by theme, publication type, author and year.
The new economic outlook for Belgium for the period 2012-2017 is based on a context of budget consolidation and weak economic growth for Europe. After a year 2012 marked by a mild recession, the euro area should gradually recover the path of growth. However, this growth would be modest and mixed according to country. The main risk factor for these growth forecasts lies in the sovereign debt crisis and the evolution of the financial sector in the euro area.
Despite this unfavourable context, the Belgium economy should avoid a recession in 2012 and register GDP growth equal to 1.4% in 2013. From 2014 onwards, Belgian economic growth should become more dynamic, without exceeding 2%. Export growth should amount to 3.7% on average on an annual basis over the period 2014-2017, which means that the loss of market share should persist (1.3 percentage points per year). Over the same period, domestic demand should have an annual growth rate of 1.6%, causing GDP to increase by 1.9% on average per year.
Belgian inflation should exceed largely 2% in 2012, owing to a new rise in energy prices, the depreciation of the euro against the dollar, and increases in indirect taxes, but should fall below 2% in 2013, notably thanks to lower oil prices. In the context of a moderate rise in international energy prices, Belgian inflation should stabilize at 1.8% on average during the period 2014-2017.
Total domestic employment should increase by 8 000 units this year and by 14 000 units next year. From 2014 onwards, total employment is expected to increase by 188 000 jobs over the period 2014-2017. The number of unemployed persons (broad administrative concept) should rise between 2012 and 2014 (+ 64 000 units). Over the following years, employment should grow more strongly, while the labour force continues to expand, partially due to the pension reform. As a result, the decrease in unemployment should remain limited to 33 000 units during the period 2015-2017. Finally, as measured by the Eurostat definition, which allows for international comparisons, the unemployment rate should amount to 7.3% in 2013, compared to 7.2% in 2011.
Driven by the federal government's consolidation measures and the federate bodies' ongoing budgetary consolidation, the general government's deficit should shrink to 2.6% of GDP this year (compared to 3.7% in 2011) and thus meet the objective of the Stability Programme. Without additional measures, the general government's deficit should again increase to 2.8% of GDP in 2013. In the medium term, the deficit should shrink slightly to attain 2.5% of GDP in 2017. To reach a balanced budget in 2015 (as planned by the Stability Programme), additional measures amounting to EUR 11 billion are thus necessary.
STU 2-12 was finalised on 1 June 2012.
The Federal Planning Bureau is responsible, within the National Accounts Institute, for producing the macroeconomic forecasts that are used to establish the federal government budget. Several hundred variables are reported in this so-called “economic budget” under the form of a single value per variable (“point estimate”), which may be considered as the best guess at the moment the forecast is elaborated. However these estimates are in fact surrounded by a sizeable degree of uncertainty. Different approaches can be applied to quantify the risks around the most likely forecast. Two of these methods are illustrated in this article using the economic budget of February 2012 as the central scenario.
Presenting forecasts under the form of point estimates offers the advantage of their being easy to understand by the public and is also required by policy makers. This is especially true when macroeconomic forecasts are established for the elaboration of the government budget, as is the case with the economic budget. Nonetheless, point estimates may give a misleading impression of precision since macroeconomic forecasts are actually surrounded by considerable uncertainty. This issue of uncertainty in forecasting can be addressed quantitatively in several ways.
A first approach consists of performing a scenario analysis to illustrate the effects of potential risks related to the baseline. Such a scenario singles out the impact of a particular shock that is considered to represent a major risk associated with the central scenario. This approach requires the use of a model, and simulation results are typically expressed as deviations from the baseline.
A second method entails the construction of confidence intervals for the forecasted variables. These confidence bands may be symmetric as is, for instance, the case with the ECB/Eurosystem staff projections, where the size of the “projection range” is set to be equal to twice the historical outlier-corrected mean absolute projection error. The range tends to widen over the projection horizon - reflecting the increased uncertainty surrounding projections when moving further into the future - and is generally larger for variables with higher intrinsic volatility. The confidence interval can also be asymmetric to reveal unbalanced risks, as is done by the Bank of England for its inflation forecasts or by the IMF for its world GDP growth prospects. Past forecast performance as well as judgment or information embedded in market indicators is used to quantify the balance of risks associated with the central scenario.
When projections are model-based, confidence intervals can also be constructed through stochastic simulations of the model, exploiting the probability distribution of the coefficients and the residuals of the econometric equations. Uncertainty associated with the assumptions of the exogenous data series can be added to the stochastic simulation process but uncertainty related to other external information, based for instance on expert opinion and introduced into the model by add-factors, is not included. For that reason, standard errors provided by stochastic simulations do not suffice for a forecast interval interpretation. Consequently, only the first two approaches will be illustrated hereafter, using the observations and forecasts in the economic budget of February 2012 as the central scenario.
Macroeconomic forecasts and analyses > Short-term forecasts and business cycle