BIMBAM

Detailed estimates of taxes for the short-term en medium term are made with the aid of the model BIMBAM. This model is the estimation tool for all taxes with the exception of wage and income tax. The latter is estimated by the model MOSI. However BIMBAM provides various inputs for this estimate. The model combines many aspects of taxes. Distinctions are made e.g. between taxes on cash basis and on EMU-accrual basis, between various types of taxpayers (households, corporations, Rest of the World) and tax receivers (state government, local government, Rest of the World). The model calculates the size of progression and the micro-tax burden. The model also shows the links between the concepts in the National accounts and those employed by the Ministry of Finance. The model also shows the impact of changes in policy on the various types of taxes. A powerpoint presentation gives an overview.

National accounts statistics and the most recent administrative information on tax revenues are used as the starting point of the estimates. National accounts statistics and administrative data on federal tax revenues/social security contributions are rather similar: taxes on a national accounts transactions basis are in the Netherlands equal to taxes on a cash basis or taxes on a cash basis with one month delay.

The method of estimation consist of three parts: a model equation, autonomous effects due changes in policy (e.g. changes in tax rates or tax deductable items) and autonomous expert adjustments, e.g. whether unexpected increases in corporation tax revenues are structural or incidental (/temporary).

For most taxes in BIMBAM, the model equation consist of a volume and price adjustment. The equations for the major types of taxes are the following. Corporation tax is explained by the change in capital income in GDP, corrected for the capital income of sectors not subject to corporation tax and fiscally compensated losses. The tax on selling dwellings is forecasted on the basis of forecasts of the construction of dwellings and housing price inflation. VAT is estimated on the basis of a breakdown of the macro-economic forecast of private final consumption (assuming no major changes in the composition and taking account of the relative importance of the different VAT-rates for each category) into nine categories (e.g. food, cars, other durables, natural gas, health care, housing services, tourist expenditure abroad) and forecasts on housing construction and other investments. Car registration tax is forecasted by a link to volume growth of GDP and a structural trendincrease in order to take account of the effect of more heavy cars. Inheritance taxes are partly linked to GDP volume growth and also to housing prices. Some excise duties are linked to GDP value growth. Energy taxes by household are linked to CPB-forecasts on the consumption of natural gas.

Changes in policy. The elaborate recording of all changes in policy and their mostly ex ante effects on tax revenues is an important part of BIMBAM. Behavioral effects are taken into account in various ways: by the macro-model SAFFIER or directly in BIMBAM on the basis of expert knowledge or models and studies (e.g. the CPB-model MIMIC or articles published); the principle is simple: we use all insights that are available and relevant (including those at the Ministry of Finance and the tax office) and make then a best guess.