Do Governments Manipulate Their Revenue Forecasts? Budget Speech and Budget Outcomes in the Canadian Provinces
This essay aims at documenting and explaining the gap between speech and action through a comparison of revenue forecasts published in Budget Speeches and actual revenues reported in provincial public accounts in Canada from 1986 to 2004. We look for two potential sources of revenue forecast errors: uncertainty and political manipulation. Our regression analysis shows that these errors are related to uncertainty: When economic conditions improve, government revenue is underestimated. Furthermore dependency on federal transfers proved to have an equivocal impact. It led to underestimation in the period of fiscal liberalism and to overestimation in the period of fiscal restraint. We also found that revenue forecasting is subject to political manipulation. Revenue is systematically overestimated in election years and governments of the right significantly underestimated their revenue in the more recent period. Finally, where there is an anti-deficit law, revenue forecast errors are lower.
KeywordsForecast Error Fiscal Policy Provincial Government Canadian Province Government Revenue
- Belongia, M.T. 1988. Are economic forecast by government agencies biased? Accurate? Federal Reserve Bank of St-Louis, November–December: 15–23.Google Scholar
- Bernard, A. 1992. Politique et gestion des finances publiques, Québec et Canada, Sillery, Presses de l’Université du Québec.Google Scholar
- Gentry, W.M. 1989. Do state revenues forecast utilize available information? National Tax Journal, 42(4): 429–439.Google Scholar
- Imbeau, Louis M. 2000. Guardians and advocates in deficit elimination: Government intervention in the budgetary process in three Canadian provinces. In Canada Observed: Perspectives from Abroad and from Within, ed. J. Kleist and S. Huffman. New York: Peter Lang.Google Scholar
- Imbeau, L.M., F. Pétry and M. Lamari. 2001. Left-right ideology and government policies: A meta analysis, European Journal of Political Research, 40: 1–29.Google Scholar