Panel data studies typically average out the error terms to be five calendar years apart such that they are less influenced by business cycle fluctuations. What are the implications of this treatment for the effects of a highly controversial forcing variable in growth models such as the ratio of government expenditures to GDP (G/Y)? We examine this issue in this paper using dynamic growth equations and we do provide two major findings. While the yearly time span is actually not prone to serial correlation problems, there is a more powerful implication: We do observe negative long-run effects of G/Y on output growth in yearly time spans, while the averaged-out panels of five-years suggest the long-run economic impact of G/Y is muted.This paper addresses this issue of ''panel averaging'' when examining panels of industrial and emerging market economies during the ''globalization years from 1986 to 2004''. With the Solow (1956) model as benchmark, the rate of population growth and the ratio of investment to output (I/Y) are the key control variables. Capital flows from abroad and government expenditures provide additional channels to the original model, through their effects to the stock of capital and to the savings rate of the economy, respectively.We feel that our research strategy is important for economic growth for several reasons. First, the optimal size of government is far from being an established issue and any positive effect on output hinges on the relative efficiency of the public sector. extends endogenous-growth models to include tax-financed government services on production and assumes exogenous government actions. find strong