Demand-side flexibility (DSF) in the electricity grid has become an active research area in recent years. While temporal flexibility (e.g. load shedding, load shifting) is already discussed intensively in literature, spatial load migration still is an under-researched type of DSF. Spatial load migration allows us to instantly migrate power consuming activities among different locations. Data centers (DCs) are power-intensive and process information goods. Since information goods are easily transferable through communication networks, power-intensive processing of information goods is not necessarily tied to a specific location. Consequently, geographically distributed DCs inherit in theory a considerable potential to globally migrate load. We analyze the economics of spatially migrating load to provide balancing power using geographically distributed DCs. We assure that neither of the participating electricity grids will be burdened by this mechanism. By using historical data to evaluate our model, we find reasonable economic incentives to migrate positive as well as negative balancing power. In addition, we find that current scenarios favor the migration of negative balancing power. Our research thus reveals realistic opportunities to virtually transfer balancing power between different market areas worldwide.