The empirical evidence of Kondratieff cycles is highly controversial. In this paper, the question is analysed how far World War I and II are responsible for this situation. It is shown that after estimating the effects of the Wars and other exogenous shocks the long-term economic development is characterized not by Kondratieff - but rather by Kuznets cycles. The long-term growth rate for the US economy seems constant, for the European economies there are significant changes in speed of the growth process. These growth phases seem to be in significant inverse correlation with the innovation process.