STU 03-05 : Special Topic - Impacts of tax shifting operations [ Short Term Update 03-05 - 05/10/2005]

Due to a deceleration in the worldwide business cycle, Belgian economic growth weakened from the fourth quarter of 2004 onwards. Although the high level of oil prices remains an important factor of uncertainty, economic activity should gain momentum during the second half of this year. As a result, GDP growth should fall from 2.6% last year to 1.4% in 2005 and to 2.2% in 2006.

This year, net exports as well as domestic demand should contribute less to economic growth than in 2004. While Belgian exports suffer from the slowdown in European economic growth, domestic demand is hampered by the poor performance of private and public consumption that is only partially compensated for by an acceleration in investment growth.

Economic growth in 2006 will mainly depend on domestic demand. Private consumption growth should increase as disposable income is boosted by the personal income tax reform and investment growth should remain strong. Combined with robust export growth, this implies an acceleration of imports, resulting in a zero contribution of net exports to economic growth.

After a net gain of 23,600 persons last year, employment is expected to record an average annual rise of 28,800 and 30,300 persons in 2005 and 2006 respectively. The unemployment rate should remain stable this year and next year.

The rise in oil prices has pushed up underlying inflation since the beginning of this year. Together with price increases in oil-related products, this should raise headline inflation from 2.1% in 2004 to 3% in 2005 and 2.9% in 2006.

For many years, the Federal Planning Bureau (FPB) has been studying the macroeconomic effects of various ways of reducing social security contributions as well as the impact of different types of alternative sources of financing. Last year, the FPB updated its usual simulation exercises at the request of the “Conseil Supérieur de l’Emploi”. Hereafter we present the main findings of these updated exercises.

Description of the simulation exercises

The short- and medium-term effects (from years t to t+7) of the considered measures have been projected using HERMES, the FPB’s macroeconometric sectoral medium- term model. All the simulation exercises have been calibrated so that the ex ante decrease or increase in government receipts is equivalent to 0.5 % of GDP over the whole period (which corresponds to 1.5 billion euro in the first year of the simulation and 1.9 billion euro in the last year).

Furthermore, these simulation exercises were carried out using the following assumption regarding wage formation 2 : the change in employment (as well as in productivity) generated by a measure has an impact on wage bargaining and therefore influences real gross wage evolution, labour supply remaining unchanged. The selected simulation exercises were as follows: Three types of measures aimed at reducing employers’ social security contributions (ESSC) were tested: a general proportional reduction in the implicit contribution rate, a reduction targeted at low-wage workers and a lump sum reduction (identical for every full-time equivalent job).

Moreover, six types of alternative financing measures were selected, which can be classified into three categories: increased indirect taxes, increased direct taxes, and new taxes. Indirect taxation was examined in terms of an increase in Value Added Tax (VAT) and an increase in excise duties on motor fuels and tobacco. Direct taxation was examined in terms of an increase in income tax and an increase in corporate tax. Finally, two new taxes calculated on very wide bases were examined:

• a “value added contribution (VAC)” based on the value added of all firms and obtained by multiplying this base with a single rate (0.83 %);
• a “general social contribution (GSC)” calculated on a very wide income base (wages, self-employed earnings, company earnings, replacement income and family benefits, capital income of the households, rents and other sources of household income) and for which the rate would be 1.17 %.
  

The results presented hereafter must be considered with caution. This warning applies in particular to the effects of the new taxes for which, by definition, many uncertainties remain. Moreover, the implementation of those new taxes could create practical or socio-political problems. It could also generate redistributive disparities. Those problems and disparities have not yet been examined, although they clearly should be before any political decision is taken. Note also that the results of all the tested measures must be considered as a general approximation of the impacts of the measures (and not as specific effects resulting from a specific way of implementing a measure).

A summary of the results of the single simulation exercises is shown in Tables 1 and 2. Table 1 displays the effects of the social security contribution reduction measures while Table 2 gives the impacts of the measures considered to compensate the decrease in receipts. The expected results of combined simulation exercises (eighteen combined exercises are theoretically possible – 3 x 6) can approximately be derived from Table 1 and Table 2. One example of results from a combined simulation is presented at the end of this special topic.

Main findings

The reduction of social security contributions targeted at low wages is the most favourable in terms of both job creation and public finances: the budgetary cost per created job would be limited to about 25,000 euro in the case of this targeted measure, whereas this cost would be, after a few years, at least twice as much in the other scenarios. However the highest GDP increase over the medium term seems to be obtained with a lump-sum reduction of ESSC. It is also worth mentioning that the option which is chosen to reduce social security contributions is not neutral in terms of sectoral impacts. [More in the publication ...]

STU 3-05 was finalised on 5 October 2005