Among the different theories of causation, there is a fairly long tradition of so-called "manipulative" approaches, which goes back to R.G. Collingwood and Douglas Gasking. Even though different manipulative accounts have diverse formulations, it is possible to isolate their common core: what makes a given event A a cause relative to a given event B is the fact that by manipulating A, namely by intervening on it and producing some changes at will, we thereby affect B and bring about changes in it. This approach has been successful especially among statisticians, economists and econometricians, and has been recently largely discussed by philosophers such as - for example - Daniel Hausman, James Woodward and Phil Dowe. This survey presents the main versions on such a perspective and their most relevant advantages and drawbacks as emerging from the current debate.