We have examined the burst lifetime distribution functions of solar wind VB s and $varepsilon$ for 1995--2005 and compared them with the same for contemporaneous SYM-H. The analysis yields clear power-law exponents of the lifetime probability distributions. When we analyzed three years around solar minimum we found similar scaling behavior for all the variables, including SYM-H. However, for the three years around solar maximum the difference between solar wind and SYM-H scaling was very clear. The fact that power laws were consistently observed for different activity thresholds show that these features are robust and repeatable, although scaling can vary depending on solar cycle. During solar maximum it appears that the scaling properties of the low-latitude magnetosphere, whose output is recorded by SYM-H, are not purely a direct response to the scale-free properties of the solar wind but are due to inherent properties of the magnetosphere. Since SYM-H scaling is remarkably robust, irrespective of solar cycle, it could be that the solar wind never acts as a direct driver for the SYM-H scaling.